Latest / Data Driven with Lou Peck / Ed Fatzinger: Deconstructing Kawasaki events, exploring Ducati DDM, and all things motorcycle data

Transcript

Ed Fatzinger: Deconstructing Kawasaki events, exploring Ducati DDM, and all things motorcycle data · Data Driven with Lou Peck

• Lou (00:00:00): Thanks again, man. I really appreciate you making the drive up here. Yeah, absolutely. Like you said, it's good that live depos aren't really a thing anymore because these treks, especially around the Los Angeles area, are not a lot of fun. Not at all. Yeah. I want to start talking about Kawasaki's, which is probably no surprise to you. My first question is, how many Kawasaki's do you have in your shop right now? Ed (00:00:23): I'm actually at a low right now. Maybe five or six, which makes the wife pretty happy. Oh yeah. Anytime a new bike rolls in, she's like, okay, what's that for? Yeah, research. It's research. Always research. Lou (00:00:37): Not a bad job to have. I got a motorcycle that's owned by the company and I get to use it for company things all the time, and it's fun. Not five or six, though. I've stuck to one kind of Swiss Army knife bike, but I had prefer to have more. Ed (00:00:54): Now that I am getting into the Ducati realm, I have my Multistrada, which is now my number one go-to bike. Lou (00:01:00): Oh, you bought Ed (00:01:01): That? Yeah, that's mine personally. And that's a glorious bike, dude. It's like a Cadillac. Yeah. I grew up more of a sport bike rider, sprinkle in with the occasional Harley Davidson, but now after riding this Multistrada, it's like a different world. Lou (00:01:18): Yeah, it's Ed (00:01:19): Wonderful. Lou (00:01:19): It's got variable damping and reactive damp, all the tools and the tricks and long suspension travel and upright ergonomics and you name it, it has it probably. Yeah. So that one you have obviously has that Bosch radar for the adaptive cruise control, Ed (00:01:38): Which is also wonderful. I thought cruise control. Why would you put that on a motorcycle? Who would ever use cruise control? I find myself using it all the time. Yeah, it's wonderful Lou (00:01:49): Highway stuff primarily, and that has saddlebags. Yes. That's what I'd love to get into some of that. My kids are, they're 12 at this point, so it's a lot of attention at home, but I'm thinking in a couple years when they don't really care about me, Ed (00:02:04): I'm just saddlebags don't fit baseball bats or Lou (00:02:06): No, none of that. Yeah, exactly. Nets and no, I can't even get my son's like, you should get a Porsche. First of all, I can't afford one. Second of all, I'm like, where are you going to sit in that and how am I going to take you to baseball practice? But I'm figuring when the kids are 15 or 16 and they don't care so much about me, then I'll just load up a Multistrada, throw some camping gear in the back and head off for a week or two. Andre Doria always texts me randomly. He's like, Hey man, I'm going to Tahoe next week for, do you want to come? And I was like, Andre, I have three 12 year olds. That's not happening. So, okay. Kawasaki, I'm going to assume 95% of people listening to this will know why Kawasaki's are cool and why you have five of them in your garage, and that's a low number for you. But for those that don't unique about the kaws Ed (00:02:58): As far as what the EDRs? Lou (00:03:00): Yeah, well, yeah. The fact that they have 'em, I guess is Ed (00:03:02): Right. So really they're the only bike out there besides the one Ducati that has 'em. Obviously it has the recording capability, and so that was sort of the interest for me was when it first came out, and I think it was Wade Bartlett, I think his article in Rider Magazine, I think it was about 2013, 2014, and that's when I first became aware of the EDR capability in these motorcycles. And so I figured, hey, I want to be at the forefront of that and do some of that research. Three weeks later, I had my little Ninja 300 motorcycle in my office and starting to tear it down. Damn. So the first thing I did once I got that little Ninja 300 was I called up Kawasaki corporate headquarters, which is right down the road here in Southern California. I said, Hey, I had this motorcycle. Can I download the EDR on it? And they're like, well, do you own it? Yeah. Is it in litigation? No. Well, then we can't download it. Whatcha talking about why does it have to be in litigation? And they said, that's just the way it is. I'm like, well, what if I was in an accident and I have no memory of it and I want to know how fast I was going. If I was going 120 miles an hour, hell, it's not going to get to litigation, so that makes no sense. (00:04:37): And they're like, no, sorry, it has to be in litigation. Well, that's when, okay, I'll do it. My damn self. Kicked in and started doing research, making phone calls, spending crazy amount of hours trying to reverse engineer and get this data and eventually got it after a good almost a year, it took me to get that. And then once we got that data, it just started snowballing from there. And Lou (00:05:13): At that point, Kawasaki was doing the downloads for free, right? Yes. I remember that. Bringing in one there, just serving popcorn. Walt Fulton was there, a a retired a a rider Hall of Famer I should say. And he was like, there's a little museum there. They're telling stories, walking you around, giving you popcorn. And then I brought another one there, mid COVID, and they had the whole place locked down just because of COVID I assume. And then the woman behind the front counter was nice enough, but she's basically made it very clear, I think it was two grand at that point maybe for the download. And she's like, we kind of wish we never started this process. We get a lot of requests. Ed (00:05:58): Well, that's when I started doing the research and started doing my own testing. I would take a few of 'em to Kawasaki every once in a while just to see what they would do on certain anomalies Lou (00:06:12): They didn't need. So it didn't need to be in litigation. You got around that? Ed (00:06:15): Well, yeah, no. So once they told me that, I then went on my own. I did my own thing, my own research, and then I caught wind of it from someone that they are doing it without litigation. And I was like, wonderful. And at this point in time, I had found a few anomalies that I just wanted to know how they would download that anomaly. So I drove down there, handed them the same thing you did, handed them the ECU and they downloaded it. And at that point in time, I wanted to start getting in their face. So every chance I got, I went down there. I'm like, Hey, how's it going? How many downloads have you been doing this month? Just trying to be super friendly. My first thought was that you were going to become the Volvo Mac of Cali. That's what one of my goals was. (00:07:07): And even one point in time, I took one down there and the same paralegal came walking out. You probably know her, I'm not going to name names. Again, friendly chatting. And all of a sudden house counsel lawyer comes walking in the front door. She's like, oh, hey, by the way, let me introduce you to Mr. House. Counsel introduced me. Sounds good. Yeah. I was like, this is the perfect day ever. This is what I wanted. And then he asked me like, Hey, do you want to go back in the conference room? You have time, you want to chat? I'm like, yeah, I'll go chat. And this is right around the time where I found an anomaly on the Ninja 400 that I was going to actually present to them and give them the information on what I found. And so we went back in the conference room, paralegal house counsel sitting there. First question out of his mouth is, why are you doing this? What are you doing? I'm like, research. I'm an accident reconstructionist. I want to know what this data is, how it triggers, educate the accident, reconstruction public. Just gather as much information as I can because quite frankly, the downloads you put out are pretty limited. Lou (00:08:25): So the first ones were not the same as you're getting today by any stretch. Ed (00:08:30): And I was going to say at this point in time, all they were giving you was four or five data plots, and that was it. You didn't even get the raw tabulated data. There's no explanation of how it triggered or what the limitations of the data were, the resolution, anything. And again, going back to our prior conversations, we had offline, we're just grateful to get this data. So any data is good data, but there has to be some sort of explanation about the data. That's kind of where I was trying to shoehorn my way in. Lou (00:09:05): You didn't get data limitations or anything like that with CDR, the first reports that I ever saw where you got the definition of each data point, but it was all in kanji and it was very limited. It wasn't even, I think it was a straight scanned PDF, the first one I got anyway, they weren't describing it. Obviously new to the forensic world as well. It seemed like this data was originally generated for maybe to prevent themselves from having product liability issues or blown engines. We see that max engine speed that the motorcycle's ever seen. So maybe it's helping them with diagnostic stuff and hey, all of our 300 engines are blowing. Well, they're all being red lined to death, so they're not used to us. Ed (00:09:56): So that was sort of my introductory statement, if you will. And then I said, if you want, we can collaborate. I can help you with a data limitation section, or I can help you validate some of these anomalies I'm seeing, or I can report anomalies that other people are seeing in the industry. We can have a relationship. And they're like, oh, well, we'll have to run it up the flagpole quote. See, because they're kind of beholden to Japan, their, I shouldn't say all back then the downloads just went to Japan to get translated and came back. So they really had nothing to do with it Lou (00:10:43): Per Ed (00:10:43): Se. I didn't know that that Lou (00:10:44): Had changed at all. Ed (00:10:45): So they were kind of like, yeah, it's a good idea, but we'll see. Kind of thing. And that was it. And then I said, oh, by the way, you have this problem with the Ninja 400, and I put together a nice little PowerPoint for 'em. Like, Hey, here's your problem with the Ninja 400. I would have one of your engineers look into it. They're like, okay, yeah, well, kind of the same thing. We'll look into it. And that was it. I didn't get any sort of response from them or I think it was more of who is this guy and why is he doing this? They were genuinely a little bit confused. To this day, I will still find anomalies here and there, nothing super major. And I'll email, I have all their emails. I'll email 'em the issue. I don't hear back from 'em. I don't get a reply, thanks or we'll look into it. It's Lou (00:11:48): Just have you seen any adjustments based on what you're telling them? In other words, you see an anomaly on the 400 and then you get another report in a few months and you're like, oh, okay. They fixed it. Ed (00:11:59): I think. So I have that belief that they're fixing it, but I don't know. I could be on the spam list and as soon as my email comes in, it's like delete. We already Lou (00:12:14): Know you're on that list for Ducati. Ed (00:12:16): Yeah, this crazy Fatzinger guy. Sorry, delete. I don't know. I like to think that they're looking at Lou (00:12:22): Them and considering them, but I don't know if I get a Kawasaki case and you're not on the other side, which is pretty rare. I'd say at this point, I'm always a little nervous if you're on the other side and I'm just relying on the Kawasaki report because there might be an anomaly that you're aware of that I'm not, and I perform my analysis. Granted, hopefully most of the time my analysis should be like, well, that's not true because of X, Y, or Z, but it may or may not happen. So I always appreciate it when you're on the plaintiff side, not that you always are by the way, but when you are, you going to go first. So I'm going to get to see your analysis and see if there's anything that contradicts what I came up with. Ed (00:12:57): Yeah, I mean, it's kind of like even on CDR reports, the steering polarity could be backwards or something, and you look at the physical evidence, you're like, I'm pretty certain my guy steered to the left, but it's showing to the right. So yeah, I mean, anomalies occur. It's going to happen. It's just how do they get addressed? I don't know. Lou (00:13:16): Yeah, yeah, that's definitely a concern. Hopefully. Well, who knows? They have product liability issues, but in dealing with motorcycle manufacturers, several of them over the years and working for them, they don't seem to get sued that often. So if they were getting sued every day, then they would probably have a really good motivation to look deeply into all this stuff and fix it right away. But maybe the demand is not there internally or the resources aren't there to devote a lot to it. Having said that, maybe they are, like you said, maybe they are taking that stuff and putting into effect, I guess time will tell. Ed (00:13:58): I think that's not getting into too far into the weeds, but on Ducati, that's probably why they put that recording capability in Ducati because it was the new tech that had the adaptive cruise control on it. So that might open up a whole can of worms litigation wise. So point they want to record some data, right? Lou (00:14:18): Yep. I mean, it's the best way to snub these cases. You and I think both know, I'm not going to speak for you. I'll just say for myself, it's extremely rare that the motorcycle causes the crash. I'm not saying it's impossible, but it's extremely rare. So if it's definitely in the best interest of the manufacturer to record as much as possible so they can just be like, I know you think it was us, but here's what actually Ed (00:14:41): Happened Lou (00:14:42): Very quickly. Same with auto manufacturers, but the EDRs and I think Wade's paper, I didn't know that Wade's paper was what first spawned your interest. That was a great paper. Wade's always so far into everything, he's the only one who could have written that paper at that Ed (00:15:02): Point. To spawn off of that, once I was able to get the data, I didn't even know how it triggered because again, there was no information from the manufacturer as to how the system worked, and there's no accelerometer on a 300. We know that. Well, yeah, I didn't even know that maybe there was at that point in time. So I just knew it involved a tip over sensor. So again, when we first were able to get the data, there was a brief moment of joy like, yes, we can get the data. But then it's like, okay, how the heck do we trigger an event? I took the tip over sensor, tipped it, shut the bike off, no event. Okay, maybe there's an accelerometer in there somewhere. I started the bike up, put the tip over, sensor back in, and literally just pushed the bike over on its side and let it fall on the floor. Which pain to me to do that? Lou (00:16:01): Yeah, your wife comes out and she's like, what are you doing with these bikes? Ed (00:16:04): Many, many handlebars later that I had to replace. It wasn't just simply falling over. So I thought, okay, maybe it's a longitudinal force. Maybe there's a longitudinal accelerometer somewhere on it. So we had these dollies at the office that we would use. They had castor wheels on it. We would use them to put the motorcycles on, move 'em around the shop, put 'em up on a pallet rack, and I thought, okay, we're going to put a little battering ram on the front of it and I'm going to hit it with one of my crash cars. I need a longitudinal force, so let's just hit it. So again, had the bike started up and I would start tapping it a little bit, put the car in park, quickly run out, tilt the tip center to make it act like it fell over nothing. I kept hitting it harder and harder, and the one time I hit it so hard, I broke all the straps and it flew off. The dolly landed on his side and truly shut off from falling again, no event. Oh, I was hoping you were going to say at that point the Lou (00:17:11): Rear wheel was spinning, so you got an event. Ed (00:17:13): So we're like, what the heck is going on? So I took a step back, I'm like, okay, what really happens in a motorcycle accident? So I brought the bike in the garage, put it on a rear wheel stand, butted the front wheel up against the wall just in case they fell off the rear wheel stand. I don't want it to go launching across into my secretary's office. And I got on it and I started riding it like I was riding on the street. I'm like, I'm like shift second gear shift, third gear. I'm driving down the road. I'm like, oh no, a car pulled out in front of me. So I nail the brakes, I stall the bike, took the tip over, sensor, tipped it as if I fell over and I got an event. It's like, awesome. Finally. So it not only triggers from the tip over, but there has to be some sort of rear wheel interaction. It's either spinning or it had been moving in the certain allotted amount of time. So that was some joy for about 15 minutes. Then I realized that there's only three events and it's done. Oh Lou (00:18:20): Yeah, I was wondering how Ed (00:18:22): You got around that Lou (00:18:23): Without buying 50 units. Ed (00:18:24): Well, that's like, oh crap. Now what do I do? So then I go on eBay, buy up all the Ninja 300 ECUs I can find. They're not cheap, man. No. A thousand dollars later, I have six ECUs or whatever it was. I download the first, had three events on it, download the second, had three events on it, download the third, had two events on it, one was open. So I bought all these ECUs and I had about three tests I could do. I'm like, well, this isn't going to fly either. (00:18:51): Then back to the drawing board research, two, three months later, we figured out how to clear it, finding a needle in a haystack in hexa ESAL data. Now you're not just reading out you need to write in. So we had to find that one value that was the EDR counter that we had to reset, and then I could do more tests. So like I said, it was a, gosh, it was almost a two year process to get to the point where I could do rapid fire testing and really dig into how this EDR system worked, and I think that's about where the 2017 SAE paper came out where I could finally get to the nuts and bolts of how it worked. Lou (00:19:41): Yeah, that's what I was wondering. You said that paper by Wade, say it's 2013, you got the 2013 ninja. The first paper is 2017. I think everybody looking on from the sidelines was like, holy cow, that's a lot of work. Trying to figure out how that all worked. And then you ended up finding out that it was for the 300, it was like a 0.6 D cell threshold prior to emergency shutdown, unless the rear wheel was still spinning at emergency shutdown. So you not only figured out what was happening, you figured out quantitatively what kind of braking rate was required to trigger the event. I always wondered about that too, and this is what I say to my class, hopefully it's not wrong. I imagine in a lot of instances, motorcycle goes into the side of a car, it's going 30 miles an hour when it hits the side of a car, and if there is only a rear wheel speed sensor, will that rear wheel experience a deceleration just be because of the impact alone that is consistent with that 0.6 Gs? So then even if it falls on its side and the rear wheel is not spinning, you still get an event. Most of the time Ed (00:20:46): It's either from the bike stalling, it'll just stop the rear wheel obviously. Or a lot of times in the impact, the rear wheel comes off the ground still spinning when it comes back down and slams, it'll stop. It'll get that sudden stop or shut the bike off. So yeah, you'll typically get that from an impact. Lou (00:21:08): And then the point, what motorcycle needed, the 0.9 Gs was that the 10 R, the ZX six ZX six needs 0.9 G. So you're not going to get that 99 times out of a hundred from actual braking from a rider. Correct. But you will potentially get that from that impact itself. Okay, that's good. So I mean, that's been my experience. I've had one case where the motorcycle, it was some dude entering a highway and he ran out of talent, put the motorcycle up against a concrete barrier, and it just kind of skidded across the barrier for a while. He unfortunately, Superman over the side and it was very high. He lived, but catastrophic injuries, unfortunately, that was the only case I've had where their algorithm did not trip an event, trigger an event, Ed (00:21:57): Did the bike fall over? Lou (00:21:58): It didn't. It just stayed. Alright, so we still downloaded it. It was such a big case, but we're like, you're not getting anything based on Ed's work. But for the most part, I've got data, almost you and I just had an inspection, I'll say recently, but it's probably within a year and you downloaded it on the spot and there was nothing in there, and it was one where we kind of expected there to be. Ed (00:22:19): Right. I've seen a lot of weird things doing the downloads over the years. I'd say my success rate as far as whether there's an event on there from an accident or not, it's probably about 50/50. Oh wow. Most of the time it's from a power failure because when the bikes fall over, it has to, first of all, that tipover sensor still has to be active and live, and then it waits two, three seconds and shuts off. It still needs power. So nine times out of out of 10, you slamming to the side of a car, some sort of power failure is going to occur. The ignition switch that hangs right up front gets blown up, so you're done. After that, you're not going to get an vent. But in our case that we had, there wasn't a lot of structural damage to it. There's power to it. It was a pretty clean motorcycle as far as accident reconstruction goes. So right. Yeah, we expect to get a download, but we didn't, the only thing I can think of was maybe the guy when he launched over the handlebars, caught the key chain, maybe turned the bike off (00:23:38): As it was tumbling, things like that. I mean, Lou (00:23:43): Yeah, so many of these cases, if you didn't have video of it or something, you'd be like, there's no way that happened. Ed (00:23:47): Yeah, it's not out of the realm of possibility. Right. And that's the only way I could explain that one. That one was kind of a weird one. Lou (00:23:56): So then you have a bachelor's in mechanical and then a master's in aero, right? Ed (00:24:02): Yes. Lou (00:24:03): I was similar. By the way, I have bachelor's in mechanical and then I was heading, I was like Air Force ROTC, looking for the arero realm. But my main point with that question is you don't have a ton of programming background. You don't have a ton of electronics background, at least formal education. I know you grew up basically hunting around a junkyard a lot, so you do have a lot of that mechanical aptitude, but that must, so was that a big hurdle that you were looking at the Kawasaki stuff and just being like, this is going to take a lot of knowledge that I don't currently have, or were you already equipped with that? Ed (00:24:39): No, I was not. I mean, that's why it took me a good two years to get it. It's a lot of research, a lot of trial and error, a lot of collaboration with other companies. Some people are like, yeah, I can get that. And other people are like, no, I can't. I dunno what you're talking about. So it was a hit or miss and it took a long time to eventually get Lou (00:24:59): It. Yeah, I can't Ed (00:25:00): Even, and I have to laugh. When I first got the Ninja 300, I called up the dealers like, Hey, I have this (00:25:08): EDR and this motorcycle, can you get this data? One guy's like, I don't know what you're talking about EDR, what is that? I'm like, okay, you clearly probably can't get it. And I called up the next dealer like, Hey, I have this in the manual. It said there's this EDR data that I can get if I bring the bike down, can you get it? He's like, yeah, bring it on down. If it's in there, I can get it. Yeah, highly doubt. You know what I'm talking about either. Yeah. So yeah, long story short, yeah, the dealers can't get it. They don't know. They probably don't even know what's in there. Lou (00:25:41): The biggest question I get a lot of the time is I get an email and it's like, Hey, how do I get this Kawi downloaded? I went to the dealer and they had no clue what I was talking about. It's like, yeah, they never do. I think there's an HQ in Canada, there's an HQ in Australia, there's an HQ in the United States. I don't know about UK, but I imagine there's somebody there that's able to do it. Ed (00:26:02): Yeah, I'd assume the headquarters of each country can handle it. Lou (00:26:07): That was one of the questions that popped up. I posted up on LinkedIn that we were going to have this conversation today and somebody asked if there is any geo luck or is it every Kawasaki around the world as far as Ed (00:26:20): It's in every Kawasaki, as far as I know. The problem is, and I've been contacted a couple times from someone from the UK like, Hey, can you download this? I'm like, yeah, I don't really have access to that ECU or that bike because it's a different part number. So I'd have to come out there and validate it before I can download your subject to ECU, which usually not going to happen to fly me out there to validate a bike. And that's the thing. My protocols that I have come from instrumented motorcycles, every single new ECU that comes out that I need to validate, I have to go out and either rent a bike or buy one. Lou (00:27:07): And renting sport bikes isn't easy. Ed (00:27:09): So there's two that I know of ridershare.com and then twistedroad.com. Lou (00:27:15): Those guys are great Rider Share. I rented one for a case recently. Oh no, it was for testing. I was doing testing on Harley Davidson, and they just rolled up with this very professional flatbed, dropped off the motorcycle. Ed (00:27:28): Oh, well, you got yours delivered. Lou (00:27:29): Have a good day. Yeah, it was great. Ed (00:27:31): Yeah. So Lou (00:27:32): What was the second one that you mentioned? Ed (00:27:34): Twisted Road. Twisted Lou (00:27:35): Road, okay. Ed (00:27:37): So yeah, I recently did a protocol on a 21, I think a 20 or 21 and up Ninja 1000, and there were none around to rent here. So I literally had to fly to Georgia, rent the bike, validate it there on the spot in some rental garage that I rented and I flew back. I mean, that's, dude, I love it. Lou (00:28:00): That's hardcore. Ed (00:28:01): Obviously I'm not going to make any money on that one. So spread out the cost. Hopefully I get a couple more for that bike, but yeah, that's what I have to do. I'm not going to just blindly download something without me validating it. Lou (00:28:16): Yeah, dude, I love that effort and you deserve all the success that comes with it because the amount of work and dedication that it takes to develop the skillset that you currently have. Yeah, I hope it's You're rewarded a lot for that. Ed (00:28:35): Thanks. And I also got another one recently where it was Ninja H2 Carbon, like they're supercharged crazy sport bike, and again, they're not around to rent in the entire country, Lou (00:28:52): Hardly around to buy, Ed (00:28:55): And there are a couple locally to buy. I'm like, that's a $35,000 motorcycle. And I told the guy, I'm like, if your client wants to put up $35,000 to buy one, I'll gladly validate it. And it's one where the ECU's blown up, so the pins are all destroyed, and so Kawasaki wouldn't be able to download something like that. They just put their hands up. Yeah, I get it. And what I could do is I could do a chip swap or a chip clone and put it on a good ECU and download it so I could still get your data, but it's not going to be cheap and I still haven't heard back from 'em. So we'll see if that ever comes to fruition. But Lou (00:29:38): One of the things I tell my clients a lot, because we end up a decent number of motorcycles because we're going to molest them in ways that doesn't feel right if we're renting. And I'll tell them, okay, the motorcycle's 10 grand, first of all, how much is at stake? How much are you being sued or suing for? And then we can sell them, say it's 10K, we can sell that motorcycle all day for 7 when we're done. So it is not actually that big of an investment. And I was at a trial recently in one of the big cross examinations for me was like, Mr. Peck, how much money did you spend? You bought a motorcycle for this. And it's like, well, first of all, yeah, that's how seriously we take assignment where we're trying to understand what happened. So we will spend a lot of money to do that. Second of all, it's not like it's my motorcycle now. My client still owns it and we're going to sell it at the end of the day. So it's ultimately like a 2 or 3 thousand hit minus whatever time it takes to buy and sell. Ed (00:30:36): I've done that on a few cases where the client bought the bike and then once the case is over, sell it for a thousand or two less and then just re-credit it back to the case. That works out. It works Lou (00:30:49): Out well. The insurance and registration sucks, but Well, that's why Ed (00:30:52): Renting's the totally ideal way to Lou (00:30:54): Go, especially with those, I don't think everybody on Rider Share. I think some of them it's like Turo style where you're just renting from Guy McFally. The motorcycle that I rented to Harley, it was clearly just a company. They own the motorcycle, they have no personal attachment to it, and I just try to sell it, tell them if I am going to do anything really funny, I'll just be like, Hey, this is what I'm going to do to the bike. Are you cool with that? If I'm just doing basic riding around and documenting, I don't do that. But I guess we should tell people too, what is on these things. So on the Kawis, it's still three events. Once it's filled, they're locked and essentially, well, how do you describe, if somebody comes up to you quickly and you have 30 seconds and you're trying to describe the whole line and maybe you don't, but what is the general triggering system for Kawasaki's? Ed (00:31:45): It boils down to the tip center. So the bike has to fall, and once it senses it's fallen, waits a few seconds and goes into what I call emergency shutdown. So once that shutdown occurs, it takes eight seconds of data and just writes it to the eprom. Some bikes require that rear wheel deceleration, some don't. So you'll need to tip over plus either the rear wheel spinning at the time of the shutdown. Other words, the bike falls, it hasn't shut down, it's just sitting there spinning at idle speed shuts down, you'll get an event. Or if it stopped, it had to have been in motion. I think it's about 10 seconds prior to shutdown. Some require a threshold of slowing down, others don't. For example, the ZX 10 and the ZX 14 and the Ninja 1000, I think you can just take your hands, roll the rear wheel just ever so slightly, tip the sensor and you'll get an event. Those don't require a sudden deceleration threshold. Why? I don't know. Lou (00:32:58): So I mean, anytime somebody tips it over in the garage, if they're, I guess that didn't happen that often with a key on, but any tip over is going to write an event, Ed (00:33:06): Right? Most of the time, yes. Lou (00:33:09): So then we get eight seconds of pre-impact, zero seconds of post Ed (00:33:14): Pre shutdown. Lou (00:33:15): Pre shutdown, right? Pre emergency shutdown, which is like 2.3 seconds post tipover. Ed (00:33:19): Right. But then you also have a problem of the tipover sensor is just a little archaic swinging pendulum. So if you knock the bike over, what if that pendulum loops a few times before it settles down? Lou (00:33:35): Interesting. Ed (00:33:35): That delays that shut down. So it has to be perfectly down for whatever it is, say three seconds and then it shuts down. If there's any sort of wiggling or moving or vibrating of that sensor and it's down, up, down, up, down, up kind of thing, it won't shut off until it's officially down three seconds and stable and shut off. Yes. Lou (00:33:59): So it's on a nice little bushing and it's relaying voltage. It's not just an on and off, it's relaying a voltage. Ed (00:34:06): Yes. So I can take that little sensor, tilt it on its side and kind of do a circular motion with the sensor and keep swinging that pendulum around and it'll never shut off even though I have the sensor on its side. Lou (00:34:20): That's interesting. Maybe that's what happened in the case that you and I were working on is either the tip over, I can't imagine this would happen very often, but maybe the tip over sensor stuck or for whatever reason it was just oscillating for, Ed (00:34:33): But Lou (00:34:33): It'll eventually, Ed (00:34:34): Yeah, it eventually stop and it'll wait that three seconds and it will shut off. So I tested that theory on my research I did with front wheel lockup and motorcycles falling. I did one on the Ninja 300 with the front wheel lockup, the bike falling, and I monitored the tip voltage as the bike sliding, and so the bike will slide in the one test I did, the bike slid for a good four or five seconds. It came to a stop, waited the three seconds, then shut off. So not only do you have the three seconds, you have to add the sliding time because during the sliding, that sensor's flipping out and spinning around. So that eight seconds that you have pre tip could very easily go off the plot or off the dataset and you won't get any pre crash data. Lou (00:35:29): So it's what we're all after is how fast is the motorcycle going Ed (00:35:33): Before Lou (00:35:34): It goes down Ed (00:35:34): So that three seconds of delay can very easily get extended. Lou (00:35:39): Yeah, so that's interesting. So if a bike, you get a left turner rider applies the front brake, locks it up, gets down to the ground slides for three or four seconds before hitting the car, then you still have this oscillation and then maybe you end up not knowing anything about pre-impact speed, that would suck. Ed (00:35:58): Yeah, I have a case right now where I can't talk too much about it, but it's an active case where motorcycle hit something on the freeway, went down at 75, 80 miles an hour and slid forever. We got the download and all you see for eight seconds is the rear wheel spinning at idle speed because it's on its side sliding. Man. So frustrating. So the pre-impact got pushed off the graph, right? Yep. And that's just gone obviously. No. So it is limited in how it triggers unfortunately, but again, it is what it is Lou (00:36:35): And we get it every time to what we were talking about earlier. We've so grateful that Kawasaki is documenting that at all and has done that. Ed (00:36:44): Absolutely. Lou (00:36:44): Every time I get a police report and I see that it's a Kawi on a new case, I'm just like, sweet. There's a chance that the traditional reconstruction methodologies for motorcycle speed determination, they're difficult to implement. They have big ranges. And if we can nail it down with the data that we get from Kawasaki, which is speed, which is coming from either the rear wheel, the front wheel, no, the rear wheel or both, that's how it is right now. Ed (00:37:15): If the bike has traction control on it, it should record front and rear wheel. Lou (00:37:21): So not even necessarily ABS. Wow. Even though they'll obviously have a channel for both wheel speeds, they don't, Ed (00:37:28): Well, you'll get a bike that doesn't have ABS, but has traction control and you'll still get front rear wheel speed, Lou (00:37:38): But not vice versa. Where if it doesn't have traction control but does have ABS or is that not Ed (00:37:45): Exist? That's a good question. No, it exists I think no, if it's say a Ninja 400, right? They don't have traction control, but if it has ABS, it has a front wheel speed and a rear wheel speed, it will still only record rear wheel speed. It won't pick up the front for what reason? I don't know. Lou (00:38:09): Which is a less reliable method for establishing the true ground speed of the bike. Right? Because it is very easy to slip the rear wheel Ed (00:38:18): Correct. And again, this goes back to Kawasaki and their downloads. They don't always give you the front wheel speed even though it's there. Lou (00:38:30): Dang. Okay, so if you get a bike that has traction, tell me if this logic is right. If you get a bike that has traffic control, it's a Kawasaki, you get a download, you get rear wheel speed only the front does exist and they really need to reach out to you to get that. Yes. Okay. Ed (00:38:44): Yeah, because I just recently was shown a Ninja 1000 download, I think it was a 2024 Ninja 1000, which is sort of the baby brother of the ZX 10 throttle by wire, pretty advanced system, traction control, power modes, rider modes, things of that nature. And the download did not have front wheel speed and on top of it, it didn't have the grip position. So if it's a throttle by wire, you'll get grip position a hundred percent, zero, a hundred percent. And throttle position you'll get both because they're not necessarily linked so well directly linked, I'll say. And in the download, you only get the throttle position again, which is unfortunate. So it'd be nice to have what the rider was intending to do versus what the bike was giving you. So you might have a road rage issue or a alleged racing issue and you get to download and you're like, oh look, the throttle position's only 50%, he wasn't racing. But if you get the grip position, you could see it pinned at a hundred percent the entire time. And it was just the ECU not giving him a hundred percent. Maybe he was trying to keep the front wheel on the ground wheelie control kicking in something and it's not giving you the a hundred percent throttle. So stuff like that can be Lou (00:40:13): Important. One of the most important things that I find in these downloads that I didn't expect was, and I think you and I worked a huge case on Sunset where this came into play, you, and I don't think either of us ever testified, maybe you did, but anyway, I did. Yeah, I remember thinking, alright, we're totally in a hundred percent alignment on what all this means and the clutch position on that Kawasaki, and I've seen in many cases, it's such a good indicator of what's going on, whether the rider pulled it in intentionally or when the motorcycle hits the car and you see the clutch lever come back. So I imagine that's some of the same might happen with the grip as well, depending on the orientation of the impact where you get some indication because finding impact isn't always Ed (00:40:59): Easy, right? It's just going to get there. So it's more like the old Ford ECMs where it's just a looping data set and there's no trigger, at least some of them. There's no trigger as to where the accident happened. You just look for the discontinuities and clutch position is certainly helpful. If you see a spike in throttle, you see the wheel deceleration and all of a sudden it drops to zero bike stalls and you see the throttle spike through the roof, well that's likely impact. Or if you have a guy that does the clutch pull during the break, so you see no clutch, then a clutch pull, then a deceleration, then a sudden drop in front wheel and the clutch release, what do you think happened or just got thrown, right? So yeah, those things are helpful in determining impact. And another thing that I noticed is kind of cool, the gear position, a lot of times you'll see the gear position switch from impact or from the bike falling down. I always knew it happened. You can never really prove it. If the bike falls on its left, there's a very good chance that it's going to shift. So people that like to find out what gear it's in, I'm like, yeah, it might not have been in that gear at the time of impact. Lou (00:42:26): You got to be careful Ed (00:42:28): And you see that a lot in the data. It's kind of cool. Yeah, Lou (00:42:31): That's interesting. So the gear position, I guess, so it's a calculation right on the Kawasaki where it's looking at engine speed and rear wheel speed and then saying, I must be in this gear Ed (00:42:41): That's only for the Ninja 300 and I think the F and the older Vulcan. Lou (00:42:48): So otherwise it's just a direct sequential understanding. Ed (00:42:52): So if your bike has a gear position indicator on the dash, in other words it tells you 1, 2, 3 that has a true gear position sensor and that will come through in the EDR data. Lou (00:43:07): That's awesome. I think that was that the same case where one of the bikes, the officer saw the tack or the instrument cluster said it was in six gear, but then they shifted and found it was in fourth. I had a case like that. I'm not sure if it was that same case. So I was doing gearing calculations just to try to corroborate my speed determination. I was like, well, it could be six, it could be fourth. We have no idea, but here's what the range suggests based on the stuck T. And that's consistent with X, Y, Z, which is one of the drums that I beat a hundred times whenever I'm talking to anybody about motorcycle recon is figure out the impact speed or the initial speed using as many methodologies as you can because none of them solely are bulletproof by themselves. None of 'em are bulletproof Ed (00:43:54): And that the whole gear calculation bikes are a whole nother can of worms. It defaults to certain values at certain times. If you get put in a false neutral, it could be show you really weird things. So yeah, be careful on those bikes. Lou (00:44:15): Yeah, I think that was another thing that I did on that case that you and I were working is I think we got front wheels and rear wheel speed. They were different but not dramatically. So it was clear that there was probably some, and the guy was breaking really well, which is some paper that I'm hoping to write with you and we've been trying to write it for five years. But I would love to write a naturalistic breaking paper based on the more I think about it, probably everything like Kawasaki EDR and video. Anytime we have a hundred percent clear evidence or 95% clear evidence of what the speed breaking rate is, because that's so limited right now. Naturalistic motorcycle braking data and that guy was breaking at like 0.85 Gs or something like that. And then the front wheel speed was a little bit, I think higher than the rear wheel, and that made total sense. But then I did a gearing calculation based on the reported gear to make sure that that aligned with all, mainly I was scared that you were going to find something that I didn't. So I was like, I'm going through everything with a fine tooth comb to make sure that when I present my analysis, ed's going to have the same answer. Ed (00:45:17): Well, that's what I do as a check on every download that I do and there is legitimate pre-impact speed, I will do that gear RPM calculation to make sure that that speeds within the certain band that I'm expecting it. Right. Lou (00:45:33): Yeah, that makes sense. And then I was thinking about that too, and I don't know how many cases I've had where this is coming into play, but it's kind of reflected in your later research, that 2018 paper where the speed that you documented via the VBOX was a few percent off of what was reported in the EDR, because of course tire size comes into play if that's what it is. It could be that. And then on a motorcycle, the tire size, the circumference is not, it's not constant. If you're leaned over at 30 degrees, you're on a different circumference than you are if you're upright. And it seems like their calculations are based on upright. So do we have to make an adjustment for that? Ed (00:46:14): Well, it's funny you bring that up because that on those earlier models, Kawasaki did chop down the speed by about 3% in their protocol. And I didn't know that at the time, and that's why when I wrote that paper, I was just taking the raw, because on speed on these motorcycles, it just goes from hex right into miles per hour or kilometers per hour, for instance. Everyone knows what FF is 2 55, right? Lou (00:46:50): Yeah. Ed (00:46:50): So it would go FF 2 55, 2 55 miles per hour. There was no conversion whatsoever. But Kawasaki would chop that down by about 3% because on some of the cases where I had their download and I had my download, I compared 'em and I'm like, oh, they're knocking it down 3% to get to, I guess they're just being conservative. They don't want to ever accuse someone of traveling too fast or Yeah. Lou (00:47:17): So that's happening in post essentially. Ed (00:47:19): Yes. And I didn't realize that, so that's why I issued a correction saying that it's not overestimated. It actually underestimated the tick based on their adjustment. Lou (00:47:30): My thought on that when I first saw it was maybe they're just accounting for the fact that, and I think my logic goes in the right direction, that 10,000 miles after the motorcycle leaves the shop, the wheel, the tire will be worn a little bit. Oh, Ed (00:47:43): Right. Yeah, that can adjust it too. Lou (00:47:45): And now they're accounting for that. So well say that the guy rides that tire for 10,000 miles and the crash happens at 5,000 miles. We're kind of split into the difference, but either way, for us 2-3% is way better than our mathematical calculations will generally Ed (00:48:02): Steer us. Right. And that's why any download that I do and give to a client, I always provide a speed validation plot right from my VBOX three I comparison. Okay, so you have the VBOX trace and you have the EDR points overlaid on it, and I always try and keep it just to tick under the VBOX to be a little conservative. Most of the time it works out to be about that 3% chop on the hex data Lou (00:48:35): You take for that process. What does that process look like? I guess you probably have two types of cases. One where somebody like me calls you and says, we have a Kawasaki, can you get involved and download it? So you're not doing any recon, basically, you're just doing the best you can to get that data and to help me validate it. And then you have a case where you guys are the recons where you probably, I don't know if you have more flexibility there, but are you getting an exemplar bike every time you get one of these cases and okay. Ed (00:49:04): Yeah. So I'd say about 80, 90% of the downloads I do are just download only. I don't do the recon. So I would download their ECU. I basically have a benchtop rig that I have for each model, say 19 to 22 ZX6. I have that wiring harness in a box, all the necessary sensors to not set additional faults. And I will put their ECU on it, download it and be done. Because I already know the protocol. I already validated it on an exemplar motorcycle. But like I said in the instance, I don't have the protocol, then I would have to go out, rent it, buy it, do whatever, and get that protocol. Lou (00:49:55): And then that's, so from your original testing of any model year bike range, you will get that validation curve that you can plot over the Ed (00:50:06): Yes. Lou (00:50:07): That's awesome. Yeah, that's good fun, man. Ed (00:50:09): Yeah, I mean, I try to give the requester as much information as I possibly can, right? Lou (00:50:19): Yeah. And that's why I wish you weren't on the other side most of the time that I was going to say, because Jarrod Carter came in here the other day, and I have such a great relationship with Jarrod and I know him so well personally in the papers he's written and all that, but I've never seen his work. Whereas you and I, you see each, at least I see you most of the time. If I'm on a case, usually it's you on the other side. Not always, it's Nathan sometimes, and there's other people obviously, but I've seen your work a lot. You happen to be on the other side of the lot. So when I have a Kawasaki, I'm pretty sure you're going to be on the other side. So I only get the fruits of your labors via discovery. But when I get the opportunity, I try to bring you on. I was talking to Eric Deyerl the other day and he's like, dude, anytime I get a Kawi call Ed right away, get him on the case and Ed (00:51:13): Make sure him and I had a power loss anomaly that we had to sort through. I'll just chat about this story real quick, where if there is any sort of power loss after the bike falls over, so you're in your accident, bike falls over, the command is sent to shut it down. Once it shuts down, it writes the data to the eprom. So then what'll happen is the ECU will sit there and wait for the key cycle to end to Lou (00:51:52): Write that data. So it's not in permanent memory until that key turns off. Ed (00:51:58): The eight seconds is Lou (00:52:01): Okay, Ed (00:52:02): But the key cycle at event, the runtime at event that is sitting there waiting for the key cycle to end. So you might have the data written to the eprom, the eight seconds, but you might not have key cycles or runtime. And so that happens in typically one of two ways. One, the bike falls over and no one ever shuts the key off and the battery runs dead or your first responders come up and clip the battery cable. So you'll have an EDR event with data, but no key cycles of runtime. So I sent that off to Kawasaki when it was free, I didn't have to pay $2,000. I said, okay, what does Kawasaki do in this situation? Will they give me the data or not? So I had one of my bikes created this anomaly, created an event, pulled the battery cable, didn't save key cycles or runtime, sent it to Kawasaki, came back with the report, they gave me the data. So I thought, cool, perfect, it's working. They still know data's in there and they'll give it to me. Fast forward about two, three years later, I find one where they don't give it to you. (00:53:30): And that was a case where I was brought on, I think it was in Kentucky somewhere where the investigating officer sent the ECU to Kawasaki. Kawasaki came back with the report and it was clearly obvious that the data wasn't from their accident. And so the one recon told the lawyer, Hey, just for the heck of it, get Fatzinger involved. Have 'em look at it just to confirm. Sent me the report, looked at it. I'm like, yeah, I don't see your data on there, but let me see the scene photographs just to confirm something, right? Sent me the scene. Photographs, obviously the front end was blown off. No headlights, no marker lights. The dash was gone. Lou (00:54:17): Not a good Ed (00:54:17): Start at face value. It looked like it wasn't on, but one of the photographs, if you zoom into the tail section, you could see the little light that lights up the license plate, that bike is still powered on. You would never know it unless you really studied the photographs. And it was one of the motorcycles where the tip sensor was pretty well protected. It was underneath the gas tank because obviously if you blow up the tip sensor, you're not going to get an vent anyway. So it had a good tip over sensor and it still had power as it was laying on the roadway. So that should have data in it. So as soon as that bike was commanded to shut down, it wrote the data and it's still sitting there waiting for the key cycle to end. And it never happened. The battery went dead. So I told the client, I'm like, I'm pretty damn certain that that's going to have data on it. Let me download it. They're like, well, they're not going to let us ship the ECU to you. It was still criminal case and they didn't want to part ways with the ECU. And I'm like, what if I fly out to download? I'm like, yeah, flew out there, downloaded it, had all the accident data right on it. No way, dude. And Kawasaki didn't give it to 'em. I'm like, oh man. So now Lou (00:55:36): I don't know, it wasn't just because it wasn't a hundred percent complete with that last little writing. Ed (00:55:42): So there was no key cycles or runtime for that event. So they just looked at that and said, oh Lou (00:55:47): Yeah, there's nothing there. Ed (00:55:49): I get it. And that's another problem with, and I hate to say problem, I don't want to trash on Kawasaki again. I love 'em, great company. So they only give you the data that they see, right? So if there's one event, they're going to give you data or plots for one event. And again, when I download it, I give you the data for all three events. So you see, okay, event number one might have real data in it, event number two or three, you'll see all Fs, F, F, F, F, F, F. So you know that there is no data there. And that's one of the pitfalls I think with Kawasaki is they see key cycles for one event, they're just going to send you one event. So there's a bit of a pitfall there. If there's a power loss, you might not see it from Kawasaki Lou (00:56:52): Similar to Kat when Jeremy Daily was developing the FLA, where it's like Kat, unless they had a full event, they're not giving you anything. And Jeremy's device could get whatever was written Ed (00:57:04): Very similar, yes. Lou (00:57:06): Alright. We had some technical difficulties but we're back on. So you were saying, we were talking at before, we had to shut down for a second, that if Kawasaki doesn't see the full event, the key cycle off, then similar to the cat, which people are probably familiar with, Kawasaki won't give you anything. Whereas your methodology, you'll at least get all the data that was written, the eight seconds. So I had to cut you off right when you were saying what you recommend the recommended process. Ed (00:57:37): Right. So if you're obviously inspecting the motorcycle, number one is find that tip center, research it prior, see where it resides in the motorcycle under the tank up front, wherever, photograph it, make sure it's still there, make sure it's still intact. So unfortunately on some of the motorcycles, for instance, say the Ninja 400, that tip sensor is literally right up in front of the motorcycle, behind the headlights, right? So if it's in a frontal accident, it gets blown up, you're not going to get any of it right? So just document that that's still there and still functioning. And then two, you don't always get this, but scene photographs, just check to see if that bike is still energized, lit up in any way, shape or form at the site or at the scene. So if you have those two things, the bike is energized and a tipover sensor is still intact. You should get data. It should be in there Lou (00:58:44): For a law enforcement officer who's first responding, a first responder you got. It's similar to an airbag control module where it's some little device on the motorcycle that has a directional arrow that says this side up, but it's just going to be a small little box, not much bigger than a couple matchboxes put together. What's your best advice for them? Identifying where the tipover sensor is and what its condition is so that they can decide whether to spend the money, which is a question mark, which we'll get back to. Ed (00:59:16): I mean the easiest way is just service manual. I YouTube, Google. I don't know. The easiest way is to get the manual. I know that's not always the cheapest way or you can ask me. I know where most of them are and I think I even have a picture on my website of what one looks like. They're all pretty similar. Yep. Lou (00:59:41): Yeah, tiny little thing. It's usually kind of on a little rubber grommet mount pointed, upright. I thought most of 'em were at the cg. I am surprised to hear that they put 'em in the front. I wonder, have you seen them evolve to keeping it away from the front? Ed (00:59:54): Nope. Yeah, some are in the tail section, some are underneath the gas tanks. Some are literally right up front behind the fairing. It's just a hit or miss Lou (01:00:04): And so Kawasaki right now, the way that I understand the process, I generally have my attorney just write to the legal counsel at Kawasaki. If we're going to be doing a download, they'll say, yep, we will be on the lookout for the package or Lou showing up and handing it off. It's a couple grand I think right now. Ed (01:00:23): Kawasaki charges $2,500 for a download and if you're law enforcement $500, Lou (01:00:29): This episode is brought to you by Lightpoint of which I am the Principal Engineer. Lightpoint provides the collision reconstruction community with data and education to facilitate and elevate analysis. Our most popular product is our exemplar vehicle point clouds. If you've ever needed to track down an exemplar, it takes hours of searching for the perfect model, awkward conversations with dealers and usually a little cache to grease the wheels then back at the office. It takes a couple more hours to stitch and clean the data. Time is the most valuable thing a person can spend. Don't waste yours doing work that's already been done. Lightpoint has a huge database of exemplar vehicles all measured with a top of the line scanner, Leica's RTC360. So no one in the community has to do it again. The exemplar point cloud is delivered in PTs format, includes the interior, and is fully cleaned and ready to drop into your favorite programs such as CloudCompare, 3DS max, Rhino, Virtual-CRASH, PC-crash, among others. (01:01:27): Head over to lightpointdata.com/datadriven to check out the database and receive 15% off your first order. That's lightpointdata.com. (01:01:34): okay, so Kawasaki is 2K. I guess you have the benefit of working directly with the manufacturer so when you go to testify, you're like, Kawasaki gave me this. The downside is that there are certain situations where they won't be giving you all the data and then they'll have to end up working with somebody like you at the end of the day anyway. And as far as I know, I think somebody emailed me the other day and they're like, Hey, Ed's already in the case. Is there anybody that else can, anybody else that can do this? But I think the answer to that is no at this point, right? I believe so, yes. Okay. So if people do want to reach out to you, by the way for that, I know you have edfatzinger.com, you're YA Engineering, they can find your information on the internet, send you an email and just say, Hey, here's the case, and then does it have to be in litigation or anything for you to do your thing? Ed (01:02:31): No. Lou (01:02:31): Okay. Ed (01:02:32): Yeah, I just have a form I send out, Hey, fill this form out, put in the proper information, send it back to me once the file is open, send it to this address and we'll get it done. Lou (01:02:44): And I imagine at this point you've testified to your own interpretations of the data many times, Ed (01:02:51): Right? Yes. Lou (01:02:52): Yeah. So it's coming in and so I shouldn't say that with a hundred percent certainty, but in my experience, every time I've seen your work, it comes in and it's obviously been validated via peer reviewed papers and everything. The thing that was interesting, I was looking at your website prepping for this and I didn't know this, but there's certain off-road Kawasaki that have data as well. Ed (01:03:14): Yes, I believe the KRX 1000, the new mule that came out, I think 24 and up, they put in the Ninja 1000 engine into it and put a new ECU in there that can be downloaded. And I think the New Ridge that came out a few years ago, that's downloadable, but they have the same sort of archaic triggering that the motorcycles have. It's just a tipover sensor. So if you're doing 60 miles an hour and drive straight into a tree and stop dead, you're not going to get a vent because the thing never rolled over. So it's again, very limited and if you are in a rollover and you roll over say three, four times and it winds up back on its wheels, guess what? It won't shut down and won't give you an event. So it's kind of bizarre. Lou (01:04:10): Yeah, that makes things difficult I guess. Not having an accelerometer on there complicates things a little bit. Ed (01:04:17): Yes. Lou (01:04:19): Alright, so shifting away from Kawasaki now the BRP stuff, Bombardier or Bombardier, I don't know how it's pronounced like that. So if the French would definitely say Bombardier recreational products, so that's like oo, Ski-Doo and then Can-Am which has a bunch of different stuff side-by-sides, the Spider, is that what that three wheel thing is called? Ed (01:04:41): Yes. Lou (01:04:43): So you've done a lot of work with that as well? Ed (01:04:46): Yes. They have what's called last instant data where typically it's just a string of 60 seconds of data from the last time the bike was running or the engine was running. It doesn't go by key cycle and go by tip over or anything. Just when the engine is running, it's recorded, so not key on No, Lou (01:05:09): That's nice. Ed (01:05:10): When the engine stops, it stops recording. It has nothing to do with the key. Obviously you need to keep started, but it triggers stop and start with the engine stop and start. The only problem with that is there's no key cycles, there's no runtime, there's no odometer stamp, there's no date GPS location, whatever. There's nothing to tie that data to your accident. It's literally just a string of data, Lou (01:05:43): Just like the old Ford PCM stuff and that's a bear man. Ed (01:05:48): And again, it's the same issue with Kawasaki and the power loss issue where if there's a power loss, your data never gets written. So if you download it after an accident, you're looking at the data, you don't know if that's from your accident or from the prior run cycle. So it's very difficult to sort of pin that data to your case. I've downloaded maybe half a dozen or so and in all the cases it was clearly obvious that the data wasn't from our accident. So it's a stretch to probably get that data. Lou (01:06:27): So I guess the situation where the data would be related is if the crash was catastrophic, engine shuts off either as a result, do they have some sort of emergency shutdown that I don't know. So the engine's got to stop for some reason or somebody's got to turn off the key even though there's carnage most likely and then it can't be restarted to get it out of that situation. Ed (01:06:53): Once you restart it, then the buffer starts recording again. Can easily override it. Lou (01:06:58): Yeah, Ed (01:06:59): But it's a nice, like you said, you can key it on all day long and you're not going to overwrite it. Lou (01:07:03): That's nice. Ed (01:07:03): It's when the engine starts. Lou (01:07:05): Yep. That's an improvement over the old Ford PCM stuff, whereas like, oh man, if that key was on until the officer came on and shut it off and then you just saw a flat line all before that, so how do we get that data? I was looking at some photographs. Steve Anderson who formerly of OPP now in the states as a private practitioner, super good at his, he does a lot of HVEDR and super smart guy, he sent over some photos and it looks like there's a little communication device that goes between the diagnostic port on most of these BRP products and then goes to your computer and then you need buds. What does that process look like? Hardware and software wise? Ed (01:07:45): Yeah, it's just the typical translator device that goes between your computer and the bike and then you need the software, the factory software, you can get versions of it online. Heck, I think you can even get it from Amazon. Lou (01:08:01): Oh, okay. Ed (01:08:02): I don't know how current it is if you get it off Amazon, but that's one avenue you can Lou (01:08:08): Search. Okay, so you can don't have to be a dealer to get the software or to get the hardware? Ed (01:08:16): Correct. Lou (01:08:16): Okay, that's nice. So then motorcycles for those that don't know, 99% of 'em have some sort of diagnostic port. You plug the tool into that and that allows you just like an OB2 port to get access to the CAN system. It appears that the BRP stuff has a very similar port. You take this little interface, hook it up to that, hook it up to your laptop use buds, which is BRP I have it written down here somewhere. Buds is BRP, Utility, and Diagnostic Software. But from what I could see, there are different levels of that and certain different levels allow you to access data that other levels don't allow you to access. So when you are buying that, you've got to get the top tier version of it and then you get 60 seconds of data similar to kind of Kawasaki stuff as far as what channels we get in, Ed (01:09:14): I think you get typical speed, RPM temperatures, things of that nature. So it's just the same general data Lou (01:09:26): And then it's like to calling back to what you're doing with Kawasaki, it's like, okay, we get that data and we have speed, but until you can testify intelligently about that, it's like where's that speed coming from? Is it an average of all four signals? What is it? We're not going to get any sort of data limitations like we do with Bosch. This stuff just gets deep quick, especially if you want to testify in an intricate case, sure can. Yes. Where you don't have other stuff to back it up, which is the off-road stuff. I get quite a few calls on off-road stuff and I, nine times out of 10 just don't do it because I have not tailored my career towards that and I feel like I might miss something that somebody who has spent their whole career doing it would see. And it's very often hard, it seems to me to get some sort of corroboration of speed. So it's like, okay, say you do download a side-by-side from BRP and you've get this speed trace that makes sense to you, but then you have no independent methodologies for establishing the speed of the thing. And then are you actually telling people the right answer? Ed (01:10:35): Off-road cases are definitely challenging for sure. Yes. Lou (01:10:38): Yeah, man. You used to dirt track, right? Ed (01:10:41): I did, yes. Lou (01:10:42): That's good fun. Ed (01:10:45): I had to stop because I would break something every year. Yeah, it got old fast. Lou (01:10:50): Yeah, that was me road racing. So then you and I have had a few emails back and forth starting to get into and dude, so many topics, we'll see what we get through, but I'm like, all right, diagnostic tools, interesting. Talk about Ducati apps, airbag vests. There's just so much out there now that might have something on it, which I think just speaks to how we need a few experts in the industry that are really focused on this stuff that everybody else can reach out to. So Ed (01:11:18): Agreed, Lou (01:11:19): People like you are crucial. But we had some emails back and forth probably a year or two ago just talking about the usefulness of diagnostic tools. In that same paper that Wade wrote in 2013 or whatever we think it was. He wrote a line in there that stuck with me, which is you might get snapshot data or something like that, but nothing useful for reconstruction. That's kind of my current philosophy on the snapshot data. It might be wrong, I'm not saying I'm not willing to change, but when you do get snapshot data, what's the latency between that diagnostic code being triggered and it being written? So if you find out that a wheel speed sensor fractures at 35 breaks at 35 miles an hour, and then that's what the snapshot data says, well how far into the impact was it? If it's a hundred milliseconds into the impact, well that wasn't the impact speed at all. So what's your current philosophy on diagnostic tools? What are you using for diagnostic tools if anything? And is that data helpful for us Recons? Ed (01:12:23): Yeah, I find that it's not terribly helpful nowadays, most of the factory software is online, so it's not a standalone program that you can get. You have to have the proper credentials to log in to even get it. So most of the manufacturer software softwares off the table. You have to go to aftermarket. I've used SNAP-on in the past. I've used platforms from texa, TEXA, but again, they don't necessarily get all the data. They'll give you the faults, but I know good and well that Kawasaki has single point snapshot data for their faults and neither of those systems I just mentioned get that data. So it's going to be a hit or miss if you even get it. And then if you get it, is it the complete record or not? Did they get all the data that's in there and then on top of it, is it even usable? Like you mentioned single point, is it just before impact? Just after impact? I don't know. Lou (01:13:39): Kind of feel like you got to do the stuff that you were doing with the Kawasaki's where it's like, okay, if I am going to use this data, then I need to go buy an exemplar, pull the wheel speed sensor and see what the timing is between the fault and the actual writing of the speed and then how often can you do that? So yeah, I cut you off there. Ed (01:13:58): Yeah, so I don't find it very helpful. 99.9% of the time. Lou (01:14:05): Yeah, and I haven't seen anybody doing the diagnostic interrogations at these high crashes even products cases for the most part. So that tells me that everybody else is in our same boat every now and again, I get an email from somebody who is like, we brought it to the dealer and we got X, Y, and Z, and it was super helpful and I'm like, dang, now these are law enforcement who can just call up Roy's towing and say, Hey, can you bring the bike over here for us? We still have the bike, it's in our control. It's a criminal case potentially Ed (01:14:37): Your best bet is dealer tools, they'll get whatever Lou (01:14:41): Is in and we can't do that. So we got to either have to hire, remember when Nissan, I don't know if this still, but you would call Nissan and they would fly a guy out for a charge obviously, but you get a Nissan person, Ed (01:14:53): You would probably have to take that bike in whatever shape or form it's in and take it to a dealer and hopefully it powers on and they can connect to it. Lou (01:15:02): My thought is like, do I need to start doing that on cases where there might be something in there that's helpful that's going so far, but Ed (01:15:12): Yeah, it's a long shot for sure. Is it worth it? Yeah, I mean it's hard to say. Lou (01:15:19): And to your point, you don't also want to get lulled into a false sense of security where you pull out your SNAP-on tool, do an interrogation, get some information either in both of our situations, I think corroborate the reconstruction with it and then not go to the dealer where they were going to get X, Y, or Z. So you don't own that. Do you still use that snap-on tool or is that something that you messed around with and were like, it's just not worth it? Ed (01:15:48): Yeah, I don't have that system anymore. Lou (01:15:50): Yeah, I keep thinking about if I had more spare time, I would mess around with it. Ed (01:15:54): I have the Texas software and I use it every chance I get just to see what I get, what the program can get out of the ECUs. Sometimes it's nothing at all. Sometimes it's snapshot data that has nothing to do with the accident, so I do use it, but like I said, 99% of the time it doesn't bear any fruit Lou (01:16:21): That's consistent with mine, so I'm glad we're in the same boat. That's consistent with my interpretation of it too. It's just like, yeah, man, I don't do it. And people, I'm sure you get this email all the time as well. Hey Ed, do you know if there's any data in a 2021 Yamaha X, Y, Z? And I'll generally respond, there's no EDR, there's no data that we know of the diagnostic tool, we don't really know. And as you start looking at some of these manuals from 2023 Yamahas plus or any motorcycle, and they'll say, Hey, we capture a lot of information. Ed (01:16:59): Yeah, they're very vague Lou (01:17:00): Sometimes They're very vague. Yeah. It's Ed (01:17:02): Like, yeah, that could kind of sound like an EDR, but not really. So you don't really know. Lou (01:17:07): Yes. Yeah. So I tell people, just read that manual, make your own decision because we haven't fleshed it out yet as a community. We don't know. Ed (01:17:14): Even the Ducati manual, they have that Ducati data monitoring, which has accident data in it, and they right in the manual said, this isn't truly an EDR, but it kind of acts as an EDR based on the description we're giving you. So they try and stay vague as possible, but it is pretty much an EDR. Lou (01:17:39): Yeah. So what'd you find out on that? How did you even get the inclination? Was it just looking through manuals and seeing DDM Ed (01:17:46): Manuals? Yeah. Lou (01:17:47): And then you're like, okay honey, I need to buy one. Ed (01:17:51): Then lo and behold, a month later, one just magically rolls into my garage and hey, let's do some research on it. Lou (01:17:56): Yeah, you can't beat that. Ed (01:17:58): Yeah, so I had to bite the bullet on that one. I bought one, and I have many stories for that system where I first bought it and once I had complete ownership of it, I had the title and everything. I went to the dealer, showed him the manual DDM Ducati data monitoring system says, owner's consent, I get the data, can you please get the data? And they're like, okay. They rolled the bike in the shop, hooked it up the computer, saw him tinkering around, then went into the manager's office, had a conversation there, a manager got on the phone, manager put the phone down, tech came back out and they're like, we can't get it. I'm like, what do you mean can't get it? They're like, well, it's not turned on in the us (01:18:47): What do you mean it's not turned on in the us? They're like, oh, it's not turned on. What are you kidding me? I just paid all this money and I can't get my own data out of the bike. And luckily I had a plan B for it, right? I did the forward or the adaptive cruise control testing on it, published on that. So I had a plan B for it. If I couldn't do the EDR stuff, I'm like, yeah, that sucks. I not turned on in the us. In hindsight, what I think they were talking about was the forward collision warning system that is not turned on in the bike in the us. So maybe they were talking about that. I don't know. Lou (01:19:26): Can the user turn that on or does that have to be No. Oh wow. Ed (01:19:29): It has the adaptive cruise control. It has the blind spot monitoring, but it doesn't have the collision warning system on it where it says flashing (01:19:39): Collision, eminent, whatever it does, that's just not turned on. It's like, okay, we will put that on the back burner. I'm going to start doing the adaptive cruise control testing. So I pull out my SNAP-on tool at the time that I had and plugged into the can, and it was monitoring all the channels and sort of validating all the can channels for when I did the testing. I went to the BBS, which is the black system on the bike. It's its own ECU if you will. It's not really an ECU, they call it. It's either BBS or BBU Black Box unit. It's basically the traction control device for the bike. And an engine has its own separate ECU. Well, that traction control unit or BBS is where that data is stored. So again, using my diagnostic tool, kind of going through all the parameters, seeing what everything had, I went to the BBS, looked at all the parameters, scroll down to the bottom. See at the very bottom it says number of EDR events. Zero. Like what? EDR events? Huh? All. Yeah, Lou (01:20:48): Dude, they're using that maybe turned on Ed (01:20:49): My bike terminology. Maybe I have that one bike in the entire US that shouldn't be here and it has it turned on. Right. How lucky can I be? Lou (01:20:56): Yeah, right. Ed (01:20:58): So at that point in time, all right, well put the bike on the center stand, turn it on, start spinning the rear wheel, took the IMU, tilted it and the Lou (01:21:11): IMU. Is Ed (01:21:11): That Lou (01:21:12): In the BBS Ed (01:21:13): Or separate? No, that's separate. Lou (01:21:14): Okay. Ed (01:21:16): Took the IMU tilted, it, simulate the bike falling over, and I did create an event like the EDR counter went to one. It's like this has it turned on. It is turned on without a doubt. So then went back to the dealer, Lou (01:21:33): Dealer, Ed (01:21:33): Second dealer, different dealer. I know I was going to get from the other one and said, Hey again, I have this DDM, can I get it? And they came back like, well, it's turned on, it's there, but we can't download it for legal reasons. Again, what are you talking about? The manual? It says it's owner's consent giving you consent, download it now we can't legally do it. It's like Lou (01:22:02): What? Ed (01:22:03): And again, in hindsight, the dealer can't do it. It has to come from corporate to allow them to get it and download it. So maybe he was saying that they couldn't do it, it has to come from corporate, I don't know. So then I took a next step and I contacted corporate and I said, Hey again, I have this DDM data I'd like to download. What do I need to do? And they came back with, this isn't really set up for consumers, it's more of a search warrant kind of thing or legal purposes. So no, we're not going to do it. I'm like, well, and then I very nicely screen cap the manual where it said owner's consent and highlighted it for 'em and sent it back to 'em. It said, no, it says third party can access it with owner's consent. I'm giving you consent, can you please download it? And then a couple emails went back and forth and they finally agreed to it. They said, it's kind of strange. You're the first person off the street. Joe below that wants to download this. So it's kind of strange why you want it. (01:23:24): I mean, all they have to do is Lou (01:23:25): Google your name, Ed (01:23:26): They'll figure it out. I have my reasons, but could you please download it? And they're like, well, we need it in writing. So I made my own little form and I filled it out and I signed it and please get the data. And I sent it to 'em and they finally agreed to it. It says, take it to XY dealer, give 'em 500 bucks. We will send them the key or the access and the software to get it. Oh damn. It's protected at the highest level in Ducati. And they downloaded it. So it's like wonderful. Okay, great. So I'm waiting for the results. I get a phone call from the technical director, north America, again, I'm not going to name names, very nice guy. He said, yeah, ed, I got your data, but I can't send you a report. Why? It's like your data is corrupt. It's like, what are you talking about? What do you mean corrupt? You didn't get it or you got it and it's just all garbage or what's the problem? (01:24:36): And he says, the first thing he got into is like, well, the mileage and the key cycles and all the runtime weren't consistent or they were off from the rest of the bike. So it was downloaded when it was on the bike, so it added access to all the other ECUs. So the mileage and the runtime deviated from the rest of the ECUs. So somehow it got corrupt. And I'm like, yeah, I can explain that. But I didn't tell 'em that because every time I would do a test, I would reflash to BBS back to its original state. Lou (01:25:15): Okay. Ed (01:25:16): So each and every test, the odometer on, say the instrument cluster would grow (01:25:22): And the odometer on the ABS kept getting reset. So they would drift apart. And he is like, yeah, for some reason the mileage and runtime and all that is corrupt. It's not matching with the rest of the ECUs. I'm like, okay. And I should preface this by the test I did was just in my garage. The bike was on a center stand, had the rear wheel spinning very similar to what I did earlier, just tilted the IMU and created an event, very simple event that I created. So yeah, he says it's corrupt in another way where we can't explain the data. I'm like, what do you mean? He's like, well, the bike shows rear wheel spinning, but no data or front wheel spinning. Front wheels completely stopped. Yeah. He's like, no one ever touches the brakes. I'm like, yeah. And then it tips over and for some reason the rear wheel is still spinning and the front wheel is still completely stopped. (01:26:28): We can't figure it out. I'm like, oh yeah, that's exactly what happened. That's not corrupt at all. Yeah, that's not corrupt. That sounds pretty normal to me. But he had no idea that I did that. And I'm like, okay, should I tell him whatever? And then he's like, and then I brought on some engineers from Italy. We got on a Zoom call and we were going over your data with the engineers from Italy trying to sort this data out. We were concerned that it was a sensor bad or was our data bad? Was something wrong on your motorcycle? And it's like, Lou (01:27:06): Oh man, I just got the whole Ducati engineering team together. Ed (01:27:10): I put my head in my hand. I'm like, oh my God, what did I just do? And I'm like, I can't not tell him now. (01:27:17): And so I told him, I'm like, listen, it wasn't an accident, it was a staged quote accident in my garage. I had the bike on a center stand, rear wheel spinning, and I tipped over to create an event. And at first he was kind of happy. He's like, oh, our bike did function normally. Nothing was bad with it. The sensors were good. And there was a long pause and then it got a little angrier and he's like, well, why the hell are you doing that again? I gave him the spiel. Hey, I'm an accident reconstructionist. This data is kind of important to our community. Just kind of doing some research, trying to figure out what data is there because I've never seen a report and I still haven't seen a report to this day of what it looks like. Oh, okay. So they never ended up giving you that report? That's good story. Dang it. (01:28:12): So he's like, yeah, that's kind of what I'm after. And he's like, it didn't really appreciate it. And then I tried to sort of make up for it. I'm like, well, I'm very thankful that you guys actually look into the data that you are that focused on having it accurate and correct and valid, unlike other manufacturers like Kawasaki. And then he got really mad. He's like, we're not Kawasaki. I'm like, yeah, I know. I'm glad you do that with your data. So yeah, the conversation didn't go very well after that. And then I asked at the very end, so you think I could get a report? Yeah, exactly. Can you still give me Lou (01:29:14): Something sitcom style a little at the while I'm leaving? Ed (01:29:18): Where's that report? Let me get with my engineers and see what I can do. And I knew as soon as I hung up the phone, I'm like, I'm not going to get anything. So yeah, wasn't a good experience. And then just to put the nail in the coffin, I did another test and I called 'em up like, Hey, I did one more test. Do you think you can download it? It's not corrupt. And they're like, yeah, we're not going to download anything from you again. Oh shit. I was like, blacklist. Lou (01:29:54): Yeah, blacklisted Ed (01:29:56): Any email from me. It goes to the Jones Box Lou (01:29:59): Standard stuff. We appreciate you getting yourself on that blacklist in the community. At least we have a better understanding of what's required. And that exists. Ed (01:30:09): Unfortunately, I'm trying to do the exact opposite. I'm trying to just understand the information so the community understands the information. I would love to work with you guys and collaborate again like I did with Kawasaki, like, Hey, I'm kind of offering my services if you want to run anything by me or do some testing or whatever. But yeah, it's note to self, if you want to work with the manufacturer, do not start reversing engineering anything from the manufacturer. Lou (01:30:44): Yeah, I guess that could piss 'em off at times. Ed (01:30:48): Yeah. So I guess it's kind of a double-edged sword. Do you do it? Do you not do it? I dunno. Just do it anyway. I know they frown upon it obviously, but it's just knowledge, right? It's a drafter. Lou (01:31:04): Yeah, exactly. And with the Kawasaki stuff, it's been really valuable. There are things that you found and you can get data now. So there's I'm sure multiple fatalities around the country right now where Kawasaki is waiting to be downloaded and the key was never turned off and never sense that cycle. And with their current protocol, Kawasaki is not going to get that data and you can. So that's really valuable and worth the effort. Ed (01:31:29): So as far as the Ducati stuff, like I said, I still haven't seen a report. I do have a couple people that have emailed me and said, Hey Ed, how did you do it? Who do I contact? What form did you use? I just send them my little dinky form and just with the caveat, Hey, if you ever get a report, lemme see what it looks like. And they're like, yeah, definitely. So yeah, I still haven't seen one. So I kind of have an idea of what's in there for my own testing. And I can tell you it's a wealth of information, like 30 40 data parameters. It's insane. You don't get as much data as far as Kawasaki has eight seconds, Ducati's maybe like four seconds, but they're a lot tighter with the triggering. It comes from a true IMU. So when it gets to a specific lean angle, it will trigger it. There's no fluctuation or vibration of anything. So it's a much tighter triggering, if you will. Lou (01:32:36): So with your methodology that you developed essentially through the Kawasaki experimentation, you're able to pull the data, you know what it looks like. You just haven't been able to validate it like you would with the Kawasaki. I mean you're monitoring the whole can traffic and looking at everything and saying, well, I have this hex data. I have this data that I know is true from this instrumentation, therefore I can make I get the Rosetta stone between the two with Ducati right now, you could probably make some very educated guesses as to what the data looks like, but not to the point where you could validate Ed (01:33:12): It. I mean, I have the main data channels, validated speed, brake pressures, clutch activation, things of that nature. But there's a ton of other data sets in there that I don't even know what the heck they are. So until I see a report, it's hard to judge what it is Lou (01:33:32): Not ready for prime time yet. And I guess if you have a case where you need to get Ducati data, you are not the point man. You got to have your attorney contact Ducati and be like, yeah. And by the way, we don't know this ED guy at all. He's not involved. Ed (01:33:45): Don't mention me. Lou (01:33:46): Yeah, exactly. You get the whole entire law firm blacklisted. That's fun. I was listening to Andre speak about this and I know you and him have done a lot of testing on airbag vests and it sounds like the communication there has been very similar where, I dunno if it was you or Andre, but it's reached out to DII or climb and been like, Hey, what do you have on your airbag vest? How do I get the data? And they're just a little bit shy, which you can understand about transferring that information to the gen pop. Where are we at on airbag vests? Ed (01:34:22): Yeah, I know Alpine Star is pretty open about their vests. If you send in a vest, they'll get you the data and I think it's only inertial type stuff. X, Y, Z, accelerations, yaw roll rates, things of that nature. But I think Andre would know the most about that. I know de they're a little tricky too. They have data. I have an inside source at Dene a A that I know there's a lot of data in there, but what they give you, the public not so much. I don't know if you subpoena the data or try and get it. I haven't ever gone to that level, but I have a Dene vest where I triggered an event and you can send it in, you can request the data and they'll get it for you. It might be three or four months until you get it back. (01:35:26): I think they ship it to Italy, and I don't know if I have this on my website, but I sent him one. I did a stage test, I put on a little dummy and put the dummy in my back of my pickup truck and drove him around town for a little while and then I stopped. I pushed him over and deployed the bag. It was pretty funny. And I was recording it, had the vbox satellite stuff all on it, and I finally got the data back. It said yes, there was a deployment. You were doing 50 miles an hour at this specific location. It had GPS position in it. The batteries were about 75% life and everything deployed. Normally that's all. It was just an email. Lou (01:36:22): They obviously looked at all the raw data and just like, here's what we're going to tell you about it. Ed (01:36:26): Right, exactly. And they didn't even tell me, okay, that the 50 or so miles per hour, is that at impact? Is that deployment or half a second prior deployment or where is that speed within your deployment? Right? Was that at algorithm enable kind of like the cars? Yep. And yeah, that was it. And I emailed them back. I'm like, Hey, I suspect there's more data in here than this. Can I get anything else? You're like, no. Lou (01:36:59): Okay. So that's what we'd probably expect on a case too. Unless like you said, unless there was Ed (01:37:03): Some sort of, so again, there's more in there, but what you can get, I don't know. But again, like you said earlier, maybe they knew who they were dealing with and they're like, quick Google search. Nope, this guy again, this is all we're going to give him. So I don't know. I know Andre sent one in and I would suspect he got it back by now. I haven't spoken to him lately, but it'd be interesting to see if he got the same result I did. Lou (01:37:35): Yeah. So that's cool. About the DII one, they got IMU plus GP PS on the vest itself. Ed (01:37:40): Absolutely. Lou (01:37:41): Dude, that's huge. The Alpine Stars one I know is I have a couple of them now and it's just IMU unless you're running the app and then it's like great, and we'll get into apps, but the problem for us as private practitioners is not getting a phone. Ed (01:37:55): Right? You very seldom do. And so even the GPS and the how accurate is it? What's the sample rate? I don't know. Lou (01:38:06): And are they calculating speed from that via some sort of doppler shift or are they just giving you absolute positions which then are not quite as useful? It sounds like there's still a lot to learn there. And then Climb came out apparently with one called the AI one, I believe that is very similar. It's got IMU and GPS. This is a joke I make during my motorcycle recon class all the time though, is the Honda Gold Wing has an airbag and it has an ACM as a result. And there's data on there potentially, but how many people on a gold wing with airbags are crashing? And it's just like not many. And I feel like that's the same way with the airbag vests. Like you and I are riding around with airbag vests. Granted, we do hooligan now and again, but we're pretty conscientious riders. Anybody who's conscientious enough to don an airbag, their likelihood of being involved in a crash that we're looking at is slim. So I haven't seen it in the real world yet, have you? Ed (01:39:05): Well, no, I haven't. And that's the thing, how many people are even wearing these vests? B? How many people even know that there's data in there? They probably get 'em cut off and thrown away by the time we get involved. So yeah, I haven't had a single case with any airbag vest. Lou (01:39:25): Yeah, me either. It's so much of the work, and I don't know why I got this random passion to do the DDA work, the Ducati data analyzer, and I was like, all right, well, this thing exists and it's a data logger that might have data. We might see it someday. I think I just kind of had a little bit of passion for it like you did with the Kawasaki. Granted, the Kawasaki's a lot more practical that data, but I just got excited about it and I was like, I'm going to go rent the Ducati and I'm going to analyze all this stuff and see what kind of data comes out of it, but it's so rare. So when I'm teaching a class for real, I just kind of gloss over real quick. I'm like, here's what to look for and here's what it could have. But the bottom line is it's so unlikely that you'll see it. Let's not spend hours talking about this. You just need to know about the fact that it might exist and how to check real quick. Whereas the Kawasaki data, it's like, no, it's really worth spending a lot of time on it because it's out there on basically every Kawasaki now. Ed (01:40:20): Well, even with the Kawasaki stuff, it took a long time to get into the mainstream. So when it first came out, it was only on two bikes in 2013 and they slowly started implementing them year by year. And it still, to this day, 10 years later, it's still sort of trickling into the industry. Not everyone knows about it. It's kind of the same thing. Bar CDR, when it first came out, it was very limited. And as more and more cars get it nowadays, that's the first thing you look at in an accident is, okay, is this car downloadable or not? And those percentages are going up because it's been around for so long, Lou (01:41:07): Which is part of why I'm hoping that this podcast, and again, thanks for taking the time, I know it's a lot. It basically blows your whole day. And I appreciate that because I think people like you and me, you're on SAE board, you're reading all these papers all the time. I'm writing my weekly email trying to find new stuff and talk to new people. And I think we get kind of this false sense that that is the entire recon community, that everybody's aware of all of this stuff. And the bottom line is they're just not. There's some people that are thrown in law enforcement into the job and they don't have a great mentor, and they're just like, this is your traffic now you're recon. And they don't necessarily have people that they can communicate a lot with. So hopefully this brings some awareness. Jarrod was in here just talking. He said he was at an SAE presentation this year, 2025 at World Congress and there was a lot of video stuff, as you probably know, some of the questions that were coming from the audience made it clear to him that people didn't understand how video worked. The fact that there are these pixels moved over from prior frames and future frames, and Ed (01:42:17): There are definitely many nuances to it. Yes, Lou (01:42:19): There's a lot to that. Ed (01:42:20): It's not just, Hey, I'm going to pick this frame and this frame and do one little calculation and I got the speed. Yeah, it might not be so accurate. Lou (01:42:27): Exactly. So in my impression with Mark Crouch everywhere and Amped's everywhere and iNPUT-ACE / Axon and these posts everywhere, I figured that it was widespread knowledge in the community at this point, which just goes to show, it's like, yeah, we've got to keep trying to spread the word on some of these nuances and Kawasaki. So I saw your post the other day on a random LinkedIn post, you replied, and it was just a guy with a Kawasaki down and he's like, investigating this crash. Here's what happened. And you're like, did you download it? I didn't see what his response was. Ed (01:43:01): I don't think he responded, but as soon as I saw that, I'm like, Hey, that's a Ninja 650. That's downloadable. And Lou (01:43:06): That's huge because who knows if that reconstructionist knew that there's data on there. And what would be somewhat tragic is if there's these crazy cases that the law enforcement are trying to figure out or private practitioners and they're just not aware that thing had data on it, that would've been the most useful information. So spreading the word and this, I'm on the tangent right now. Go for it. Bottom line is look at the manual. So go to Kawasaki, look up the whatever model year you're interested in, go to the manual and try to find the term EDR. Right. Ed (01:43:44): That's the easiest way. Yeah. Lou (01:43:45): Okay. Except for the dirt stuff. You might have to be a little more nuanced. Ed (01:43:49): Well, no, they're in the manuals now for the off-road stuff, when the K RX 1001st came out with the capability, it was not in the manual, and it's still not to this day. I think on the 2020 and the 2021 model, it's not in there. And I don't know if they put it in there because I had mentioned something about it or they got wind of it because all my conferences, I was like, Hey, some side-by-sides or some off-road vehicles may have it, and it may not be in the manual. So if you're unsure, just contact Kawasaki and say, Hey, I have this model. I'm going to send you to ECU to download it and see what they say. I mean, they might have more information that we don't know. Not in the manuals. And so now 2023 and up, it's in the manuals. I don't know why they didn't go back and put it into 2021. Maybe it's already out in print. They didn't want to change it, but yeah, so they didn't do it at first. Lou (01:44:58): Okay. So yeah, hopefully, I mean, if that's the only thing that people pull out of this podcast, I think it'll be useful is if you got a Kawasaki, it probably has data, check the owner's manuals and then go from there, contact ed for a download, go to Kawasaki for a download. Ed (01:45:14): Yeah, I try and keep my website pretty current on the year, make and model sort of list. Lou (01:45:19): Yeah, that's right. That's huge. That's worth mentioning. If you go, it's right on your homepage, I think. Or maybe if you go to ka, scroll down a little bit, there'll be a table showing every year that the motorcycle has been covered in green. Alright. So yeah, the miscellaneous stuff, I didn't know, so I can't remember how I found out about, oh, I had a case. It's settled now. It was out in Arizona. I don't think you were on the other side of that. No. Steve Anderson was on the other side of that, and there was a photograph from one of the scene officers of what appeared to be a GoPro on the handlebars. At first I thought it was a GoPro or something. Then I was like, no, I think that's GPS navigation. And then Sam pulled some magic and did some Google image searching. (01:46:01): It's like a nighttime, so you don't get a lot of detail. But I found out that it was a Power vision, power Vision five or something like that, specifically for the Harley Davidson. And it just plugs into the diagnostic port. And then the rider, it's again, one of these data loggers. So the rider has to press record, but when they do press record, you get everything and it's glorious. It's like 10, 20, 30 hertz depending, and it tells you the millisecond it was written and it's everything. Unfortunately, that motorcycle was long tossed by the time I got involved in the case, but I was like, shoot, okay, well, that's something I got to keep an eye out for because Ed (01:46:34): Well, you have a perfect example when you teach your class, always pull seats, always look for any sort of device that doesn't belong there or doesn't appear to belong there. You never know. You never know what the heck's in and pull side covers. I've seen instruments stuffed in side covers of air boxes. You never know. Lou (01:46:58): That was my biggest takeaway from, I think Jeff Bonsal was there and I was at NAPARS and I was given a presentation on here's the current state of motorcycle, EDR, here's everything we know. And I was like, Ed, are you here? Is Ed here? And then Jeff's like, I got him on the speakerphone right now. But that was my biggest takeaway from doing the research necessary to make that presentation, which is like, look at the diagnostic port of the motorcycle. Do not leave that motorcycle without putting your eyes on the diagnostic port. Because Tony Andreoni, during my presentation at NAPARS, he's like, Hey Lou, you ever seen the Screaming Eagle Pro Street tuner? And I was like, no. And it's like this big, it's the size again of a couple matchbooks, and it plugs directly into the diagnostic port on Harley's and it records everything apparently. Oh, wow. And I was like, you got to be kidding me. So have I ever had a case where that thing was there and I just overlooked it? Ed (01:47:55): I hope not. Yeah. We had a case a while ago where there was a device in the tail section of the seat. It had a couple serial numbers on it, but no real description or no logo on it or anything. And so we photographed it. We always do and do some research later. But another expert, he was a trucking expert, saw my inspection photographs. He's like, I know what that is. That's a GPS tracker. We use that on our fleets all the time. I'm like, no way. Lou (01:48:27): Stolen vehicle type of a thing. Yeah, Ed (01:48:28): Exactly. And the kid just bought the bikes. You had no clue it was on there. So it was actually registered to the prior owner? Subpoenaed? The prior owners records. Got it. It was still active. We knew exactly how fast he was going prior to kidding. Yeah, the case settled in a day. Lou (01:48:45): So it wasn't GPS once a minute. It was, oh, it is Ed (01:48:49): One Hertz data. Holy Lou (01:48:50): Cow. Nice clean GPS data. Dang, I had a case like that. I can't imagine it was the same case. I would've probably seen your data, but it same exact thing happened. I'm photographing under the tail trying to do my due diligence. I find this weird device. I come back to the office and I Google the crap out of it and find out that it's like a low jack type of a thing. That case ended up settling before we ended up getting the data, unfortunately. But that's a good point. It's like, man, but if you're not, you're not inspecting motorcycles every day or very often. Some of that stuff is hard to spot Ed (01:49:25): Or there's no key. How do you take the seats off? So sometimes you're hamstrung, but yeah, it all depends. Lou (01:49:35): So the power vision stuff, which is by Power Commander for the longest time I figured any Power Commander thing essentially is if I see Power Commander, it's a non-issue to me. Somebody trying to control their fuel, smooth out the throttle, maybe get some more horsepower, but there's no data on there for me. Then Power Vision changed that for me, and it's like there's the power Vision three, four, and five. One of 'em just plugs into the diagnostic port and to your point is going to just be behind this plastic panel and you're not going to have any other evidence of it. So if you don't take off plastic panels, you won't see it. Other ones like Handlebar Mount and things like that, the Power Commander for Power Vision, I'm sorry, four and five Power commanders themselves have no data on them. Have you ever seen, so the wool or is it Woolwich? Woolwich was very important to your Kawasaki research. I suspect it's important to your Ducati research. That is a data logger that can get an immense amount of data, and essentially if it is installed, it will be recording. Have you seen that thing in the wild? Never. Ed (01:50:39): Yeah. Yeah. I just bought that for the initial Kawasaki research. I didn't use it on the Ducati stuff, but yeah, it's very similar. You plug it in and those devices, as soon as you turn the bike on, it senses that power and starts recording immediately Lou (01:50:57): To a microSD card, which is like gold. Just easy. We don't have to involve any Italians or Ed (01:51:03): Anything, unlike the phone apps, which we'll probably get into. You have to manually trigger most of 'em, and it's just how often are you going to see that? Lou (01:51:13): Yeah. Yeah. I'm looking forward to talking about that. There was a couple things, really, the Kawasaki rideology and then whatever the BMW one is. I know Eric Deyerl l did a lot of research on that, and that was interesting. So the DA, we talked about that worth knowing about. What about the Yamaha CCU, which is built into the bike, but it's kind of a data logger thing that you access with an app when you're close to. Have you ever seen anything? Ed (01:51:39): No. No. Again, it's few and far between, whether it's triggered or not, who knows? Lou (01:51:46): Yeah, same here. I'm in the same boat. I just wanted to make sure. I never want to be missing anything. I don't want anybody in the recon community to be missing anything. Ed (01:51:54): Yeah, I mean, you bring up a good point. Go straight to the data diagnostic connector. Is there anything connected? Yeah, Lou (01:52:01): That's huge. And I wouldn't have known that until really recently. So then you have all the apps. Some of them, I guess the first time that I saw an app where I was like, oh man, I really need to look further into this, was the REVER app, R-E-V-E-R, which I try to do a little bit of research to figure out how many downloads I had. It's not public information, but it's got a ton of reviews, like thousands and thousands of reviews. I had a case where a rider was using it, and he actually produced printouts from it, which had color coded segments of GPS with respect to his speed. It unfortunately got blown up because he flew away in a helicopter, unfortunately. So the speed went up to 180, so the gradient, it ruined the resolution of the gradient. It looked like he was going maybe 35 based on that, but it was hard to tell. (01:52:57): I used Photoshop and tried to get a color matcher, and I'm not sure how fruitful that was, but he provided the data and it had an elevation plot, which was all beautiful, and I was like, oh, the speed plot's going to be great. There's no speed. I couldn't find any speed. So the best you could do is get access to their dashboard and then figure out exactly what second, they're at each GPS spot, but nothing super useful. But the apps that seem, then the rest of the manufacturers seem to have apps that are just like, okay, if you want to record your ride, hit record and have fun. Ed (01:53:34): Right. Unfortunately, it's still manually activated, except the only one I know of is BMW, where it automatically triggers. You could set it up to once your phone pairs or the bike starts up, it'll start recording, which is nice. All the other apps, I don't think they do that, but I don't keep track of all their updates. The radiology one for Kawasaki, the original one that I had at the start of my research, you couldn't export any of the data. The second version came out, you can export the data. A third version just came out. I checked before coming up here. A new version came out in February. It's a nice new clean version, but I don't think there's anything too different on it. So it's a matter of just staying up on the updates or stay current on the updates. But like you said earlier, are we even going to get access to this? Because on their phone, how often do we get phones to interrogate? Not very often. Lou (01:54:41): No. It's so rare, especially if you're hired a couple years after, it's like they would have to save it or you'd have to get access to their dashboard. I guess that'd be one way. Rev had a dashboard, so that case ended up settling out, but I told the attorney try to get login information to their dashboard if that's legally applicable, because then I can get all of the raw data. But if it's being stored locally on a phone and it's been two years and then they've got a new phone, it seems like it's going to be very unlikely. So the radiology stuff, which is the app that you've done the most work on, there's a couple interesting things there. One, some of these are questions, but you have to record and then you get GPS in the phone, but it's also connecting to the bike and getting telemetry from the motorcycle itself. Ed (01:55:33): Correct? Right. Lou (01:55:34): That's awesome. Ed (01:55:35): Yeah, I believe it's at one Hertz, it's a pretty good sample rate. Lou (01:55:40): So the rider still has to hit record. It's not going to do the BMW type thing? Correct. And then does that get uploaded to a cloud or is it still stored locally? Ed (01:55:51): I think it's locally to the phone. I don't think it's a cloud. Okay. I don't remember having to create an account or log in. It's funny you asked. When I updated the app this morning, it still had my old rides from when I was doing the research on it four years ago. Oh, okay. I still have the same phone and they're still on it, so they don't go away. They don't get overridden. It's still on your phone if you have the phone. So say obviously you can get GPS data from your phone, just a straight phone. It records a certain amount of data, but that gets erased very quickly over time, weeks, months, whatever. But if you record a trip in ology, it stays there until you physically delete it. But I don't think it goes to a cloud like you were suggesting. Lou (01:56:48): That's handy. So as long as, I guess that's probably most useful for law enforcement officers, and I was talking to several of them, and it sounded like it really depends on the jurisdiction as to whether or not they have the warrant power to get the person's phone. And it's like, man, if they can't even get the phone, all we're going to do is an interrogatories. Be like, can we please have your ology data? And if the guy's going 130, what Ed (01:57:16): Data Lou (01:57:16): Is going to give that to you? Exactly. What are the odds of that? So I guess it's most useful on the law enforcement side if they know about that and if they do have the legal power to get the person's phone or to access their cloud data. But it sounds like they do need the phone, then that's useful, but it's like, all right, good to know about. Even if you're just asking the question at Depo, that's great. You tell your attorney, Hey, the guy's riding a Ducati, ask him if he had the Ducati Link app or whatever it's called, and if he was running it. I had a case recently where that happened. It was a Yamaha infotainment system suggested that if the rider is navigating at the time, then there might be data. So I was just like, Hey, ask him. And he's like, yeah, no, I was running it. He's like, okay, he just got navigation. He's like, yeah, so yeah, come on up, do a download. So it's great. It's just like, that's how I feel about the boombox as well. The boombox is like, dude, the odds of you having the rider recording the rider are so slim, but don't show up to trial and have that cross-examination question be available. Ed (01:58:21): Why did you look into it? Right. Lou (01:58:22): Yeah. My biggest thing is maintaining credibility and letting the jury know that I turned over every stone. I don't want to come up looking like some lazy schmuck, and I try hard not to make that happen. But yeah. I think we covered all the data. Is there any other data that I haven't brought up that exists? Ed (01:58:46): I don't think so. Lou (01:58:51): I told you it'd be two hours, but then we had some technical difficulty, so it's like one o'clock. So it's been about three hours. Again, I really appreciate you coming up and spending the time and helping to spread the information. If people want to keep up to date with you, it seems like the best place is LinkedIn and then edfatzinger.com. Ed (01:59:08): Sure. Yeah. Lou (01:59:11): Yeah, man. Ed (01:59:11): But yeah, appreciate you having me up. Lou (01:59:13): Yeah, it's been fun. Hey everyone, one more thing before you get back to business, and that is my weekly bite-sized email. To the Point would you like to get an email from me every Friday discussing a single tool, paper method, or update in the community. Past topics have covered Toyota's vehicle control history, including a coverage chart, aADAS, that's advanced driver assistance systems, Tesla vehicle data reports, free video analysis tools and handheld scanners. If that sounds enjoyable and useful, head to lightpointdata.com/tothepoint to get the very next one.