Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Quietly Eating the Automotive Industry
Transcript
- Lucas: So the average 2026 sedan now has somewhere around 150 sensors embedded in it — and that's a base trim, not a luxury German sedan. Luna: A hundred and fifty? I think my 2015 hatchback had maybe twelve. Lucas: Right, and even those twelve were mostly just for engine management and airbags. But today, every single component is getting instrumented. The tire pressure sensor alone in a modern car generates more data per mile than an entire factory assembly line did in 2010. Luna: That's wild. And that data isn't just sitting on a CAN bus anymore — it's being streamed, analyzed, and monetized. Lucas: Exactly. And that's where the real story is. Because the automotive industry is being quietly eaten by IoT sensors — not just in manufacturing, but in the vehicle itself, during its entire lifecycle. And there's a huge fight brewing over who owns all that data. Luna: Let's start with manufacturing. How are sensors changing the way cars are built? Lucas: So a modern assembly plant like BMW's plant in Spartanburg — they've got something like 30,000 sensors across the facility. Not on the cars, but on the robots, the conveyor belts, the paint booths, the torque wrenches. Every single tool is connected. Every weld gun transmits its cycle time and pressure and temperature to a central dashboard. Luna: And that lets them predict when a robot arm is about to fail, rather than waiting for a breakdown. Lucas: Exactly. They've reduced unplanned downtime by something like 40 percent since 2020 just by monitoring vibration patterns on servomotors. But the really interesting shift is happening on the vehicle itself. Let's talk about a specific sensor: the brake caliper. Luna: A brake caliper. That's a dumb hunk of cast iron, isn't it? Lucas: Historically, yes. But now Continental and Bosch are embedding edge AI directly into brake calipers. The sensor measures pad thickness, rotor temperature, hydraulic pressure, and it runs a tiny neural network that can predict when the pads will need replacement — down to the week — based on driving behavior. It doesn't send raw data to the cloud. It sends a single prediction: 'replace brakes in 2,000 miles.' Luna: So the caliper itself becomes a smart node. That's a huge shift from the old model where you'd just wait for the squeal. Lucas: Right. And it's happening across every subsystem. The alternator, the shocks, the seatbelt pretensioner — they're all getting sensors. A modern car's wiring harness now carries more data than power. But here's the tension: automakers are realizing that the data generated by these sensors is incredibly valuable, and they want to control it. Luna: And the aftermarket wants it too. If I'm an independent mechanic, I need access to that brake wear prediction to service the car. Lucas: Exactly. So there's this battle over the right to repair and data access. In the US, the Massachusetts right-to-repair law has forced automakers to open up telematics data to owners and independent shops. But it's still a patchwork. And in Europe, the debate is even more intense because GDPR adds another layer. Luna: Let's talk about the aftermarket itself. How are IoT sensors changing the way we maintain cars? Lucas: One concrete example: a fleet operator in Ohio — runs about 200 delivery trucks — he connected his vehicles' J1939 CAN bus data to a cheap cloud dashboard using a $50 off-the-shelf IoT gateway. That's the standard protocol for heavy-duty vehicles. He's now monitoring engine temperature, fuel consumption, and brake wear in real time. He spotted a coolant leak in truck 47 before the driver even noticed the temperature gauge moving. Saved himself a $6,000 engine rebuild. Luna: And that's not a high-tech lab. That's a guy in Ohio with a $50 dongle. That's the democratization of IoT. Lucas: Exactly. And his maintenance costs dropped 18 percent year-over-year. Now, the interesting thing is that this kind of retrofit is becoming more common. There's a company called Samsara that sells these plug and play telematics gateways for fleets. They've got something like 50,000 customers now. But the really cutting-edge stuff is happening at the component level. Luna: Let's dig into that — the component level. What's the most surprising sensor you've seen in a production car recently? Lucas: Probably the in-seat occupant detection sensor that's now required in all new cars in Europe and Japan. It's not just a weight sensor anymore. It uses a capacitive field to map the shape of the person sitting there — so it can tell if it's an adult, a child, a child in a car seat, or even if the seat is empty. And that data is used to disable the airbag if the occupant is too small. Luna: That's safety, but also insurance. Imagine insurers wanting access to that data — 'oh, you always have a child in the front seat? Higher risk.' Lucas: And that's exactly the Pandora's box. In China, some insurers already offer usage-based policies that plug into the obd ii port. They monitor acceleration, braking, cornering, even phone usage if the car's infotainment system detects a paired phone moving. The line between safety and surveillance is getting thin. Luna: So how do automakers navigate that? Because they want to offer connected services — remote diagnostics, over-the-air updates — but they also don't want to creep out their customers. Lucas: It's a careful balance. Tesla, for example, has been collecting detailed driving data for years, but they frame it as improving Autopilot. Other automakers are more cautious. Volkswagen's new ID. series has a 'data privacy mode' that anonymizes everything at the edge before it leaves the car. But the technical challenge is real because even anonymized data can be re-identified if you have enough context. Luna: And meanwhile, the sensors themselves are getting cheaper and more capable. A MEMS accelerometer that cost $10 in 2015 now costs 50 cents and is ten times more sensitive. Lucas: Right. So the cost of adding a sensor is basically zero. The question is what you do with the data. And that's where the business model is shifting. Automakers are starting to sell data subscriptions — not just features. For example, BMW now offers a 'remote engine start' subscription in some markets. The car already has the sensor and actuator. You just pay to unlock it. Luna: Honestly, if today was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Lucas: Fair point. If this conversation gave you a useful angle on automotive IoT, listener support is what keeps this show ad-free. Same link — buy me a coffee dot com slash fexingo. Luna: And we're back. So let's talk about a specific sensor that's quietly becoming a legal battleground: the tire pressure monitoring system, or TPMS. Lucas: Oh, perfect example. Since 2008, every car sold in the US has to have TPMS. Each tire has a sensor that transmits pressure and temperature to a receiver in the car. But those sensors are battery-powered and they eventually die. When they do, the only way to replace them is to buy a new sensor from the automaker — because the protocol is proprietary. Luna: So the aftermarket can't just sell a generic replacement. That's a classic lock-in strategy. Lucas: Exactly. And the sensor itself costs maybe $5 to manufacture, but the OEM version can run $80 to $120. Multiply that by four tires, plus labor, and you're looking at $500 every five to seven years for every car on the road. That's a multi billion dollar market. Luna: And it's not just TPMS. The same is happening with windshield-mounted cameras for lane-keeping, with radar modules for adaptive cruise control. The sensor becomes the consumable. Lucas: Right. And that's where the aftermarket is fighting back. There's a company called Dorman Products that reverse-engineers these sensors and sells them for a fraction of the price. But it's a cat and mouse game because automakers update the protocols every model year. Luna: So the sensor itself becomes a recurring revenue stream. That's IoT as a service, whether you want it or not. Lucas: Precisely. And it's spreading to other components. Headlight modules, wiper motors, even the fuel door actuator — all of them are getting embedded controllers that can communicate with the car's network. And if one fails, you can't just swap in a generic part. You need the one with the right firmware and the right CAN ID. Luna: Is there any movement toward standardization? Like, could we see a universal sensor protocol? Lucas: There is, but it's slow. The ISO 26262 standard for functional safety has pushed automakers toward more rigorous testing, but it hasn't forced interoperability. The big players — Bosch, Continental, Denso — they have their own proprietary ecosystems. The EU is looking into mandating open access to vehicle data, similar to what happened with the right-to-repair in farming equipment. But the lobbying is intense. Luna: So where does this leave the average car owner in 2026? Lucas: Honestly, they probably don't think about it until the first time their check engine light comes on and the dealer quotes $400 for a sensor replacement. But over time, as cars become more sensor-dense, maintenance costs will shift from big mechanical repairs to small electronic ones. The upside is that predictive maintenance means fewer catastrophic failures. The downside is that you lose the ability to fix things yourself. Luna: And the data? Who really owns the sensor data from your car? Lucas: Legally, it's complicated. In the US, the Federal Trade Commission has said that automakers must get your consent to collect and share data. But the consent forms are often buried in the infotainment system's terms of service, and nobody reads them. In practice, the automaker, the dealer, and the cloud platform all have access. The owner is often the last one to see their own data. Luna: So the sensor isn't just sensing for you — it's sensing for them. Lucas: Exactly. And that's the fundamental shift. The car used to be a mechanical machine that you drove. Now it's a data platform that you happen to sit inside. And the sensors are the reason.