Latest / The Manufacturing Podcast with Fexingo: Factories, Supply Chains, and Industrial Business / Why Factory Robots Are Stuck In The Middle
Transcript
- Lucas: The thing about factory floors in late twenty twenty six is that the big news isn't about replacing humans anymore. It's about the awkward middle ground where machines and people actually have to work side by side without getting hurt or going insane. Luna: Right, because for a decade it was all about cages and fences. Now everyone wants the speed of automation but keeps their workforce right next to it. Lucas: Exactly. And that shift has created this massive market for what the industry calls collaborative robots, or cobots. They aren't just arms in cages anymore. They are lighter, they have force sensors built in, and they stop instantly if they bump into a person. But here is the catch I want to drill into today: the hardware is easy. The integration is a nightmare. Luna: So you are saying the robots themselves are fine, but getting them to talk to the existing legacy equipment is where the money disappears? Lucas: Precisely. Let me give you a concrete example. There is a mid-sized packaging manufacturer in Ohio that I visited last month. They installed six cobot units to handle end of line palletizing. The robots cost about thirty thousand dollars each. Very reasonable. But the engineering firm they hired to integrate those robots with their old conveyor belts charged nearly double that amount. Luna: Double the hardware cost just for setup? That seems extreme for something that should be plug and play. Lucas: It is extreme, but it makes sense when you look at the reality of older factories. Most of these facilities were built before cobots existed. Their PLCs, their programmable logic controllers, are fifteen years old. They don't have standard communication protocols like OPC UA. So every single cobot needs a custom interface box, a dedicated engineer to write the handshake code, and months of testing to ensure it doesn't crash the main line. Luna: That explains why the ROI models always look so optimistic until the first invoice arrives. You are betting on cheap hardware but paying premium labor to make it work. Lucas: And that is the paradox. Cobots are marketed as democratizing automation for small businesses. But if your total cost of ownership includes that expensive integration layer, you are no longer cheaper than buying one big traditional robot arm that does everything automatically, even if it takes up more floor space. Luna: So who is actually buying these? Is it really the small shop, or is it larger companies trying to solve a specific bottleneck? Lucas: It turns out it is mostly the larger companies, but for a very specific reason. They are using cobots to fix the changeover problem. Remember episode two twenty two on changeovers? This is the solution to that. A large automotive parts supplier uses cobots because they can be moved around the factory floor in an afternoon. A traditional cell requires weeks of retooling. If demand shifts from truck parts to sedan parts, they just roll the cobot over there and reprogram it in an hour. Luna: That mobility is huge. It effectively turns fixed capital into variable capacity. You aren't locked into one product forever. Lucas: Yes, but you still have that integration drag. I spoke with the plant manager there, and he admitted that while the cobots save them from hiring temporary workers during peak season, the maintenance burden is higher. Every time you move a cobot, you have to recalibrate its sensors and check its safety zones. It is not truly zero-touch. Luna: Recalibration sounds like a hidden cost that nobody puts in the pitch deck. Do these robots actually require less skilled labor to operate, or do they just require different skills? Lucas: Different skills, definitely. The old model was that you hire a robot technician who knows how to fix the mechanical arm. The new model is that you need someone who understands basic programming logic and safety compliance. It is a slightly higher bar than pushing a button, but much lower than being a robotics engineer. It bridges the gap nicely. Luna: So it helps with the labor shortage we talked about in episode two zero seven, by making the job accessible to more people? Lucas: It does. And that is the second angle I want to explore. Safety culture. Because these robots work near humans, the training requirement changes completely. You cannot just tell a worker 'stay behind the yellow line.' They are inside the zone. So companies have to invest heavily in training programs to teach operators how to interact with the machine safely. It creates a more engaged workforce, but it also slows down the initial ramp-up period. Luna: I would imagine that slowing down the ramp-up is a dealbreaker for some CEOs who just want the output yesterday. How do they justify that delay? Lucas: They justify it by looking at error rates. Traditional automation is fast but brittle. If a part is slightly misaligned, the robot jams, stops the whole line, and requires a reset. Cobots are softer. They can handle minor variations in input. They adapt to the part rather than forcing the part to be perfect. For low volume, high mix production, that flexibility beats raw speed every time. Luna: High mix, low volume. That is the sweet spot for these things. It makes sense that mass producers are still sticking to the big cages. Lucas: Correct. If you are making ten million identical widgets, you do not want a cobot. You want a stamping press that runs twenty four hours without stopping. But if you are making five hundred custom medical devices a day, the cobot is the only way to go. The economics flip entirely based on variety. Luna: So the lesson here is that technology is not one size fits all. You have to match the robot to the production profile, not just the budget. Lucas: Exactly. And that matching process is where most managers fail. They buy the robot because it looks cool on a trade show floor, but they ignore the production profile. They try to use a cobot for a high-speed task it wasn't designed for, and then they wonder why the ROI is negative. Luna: That sounds like a classic trap. Buying the shiny object instead of solving the actual operational constraint. Lucas: It is. And I think we see this across many industries now. It is not just packaging or auto parts. Food processing is starting to adopt cobots for delicate tasks like sorting fruit or packing baked goods. The sensitivity required there means you cannot use a rigid arm. You need the force feedback that cobots provide. Luna: Sorting fruit is a great example. Humans are good at it, but they get tired. Machines are consistent but too rough. The cobot sits right in the middle. Lucas: And that middle ground is expanding rapidly. We are seeing software platforms emerge that simplify the integration piece. Instead of writing custom code for every old PLC, these new middleware solutions act as translators. They sit between the legacy machine and the new robot and handle the data exchange automatically. Luna: Middleware. That sounds like another cost center though. Are these platforms affordable for the average factory? Lucas: They are subscription based, usually a few thousand dollars a year per cell. Compared to the tens of thousands in engineering fees, it is a bargain. But it introduces a recurring cost that managers didn't have before. It shifts the expense from capital expenditure to operating expenditure, which actually makes it easier to approve in today's budget climate. Luna: CapEx to OpEx shift. That is a powerful financial lever. It changes the conversation from 'can we afford this machine' to 'can we afford this monthly service fee'. Lucas: Exactly. And for smaller firms, that liquidity matters. They don't have cash reserves to drop fifty thousand dollars on a project. But they can handle a monthly subscription. It lowers the barrier to entry for automation significantly. Luna: So we are moving toward a world where automation is more like software than hardware. You subscribe to the capability rather than owning the asset outright. Lucas: I think that is the inevitable direction. Even the hardware vendors are starting to offer leasing models. You pay for the uptime, not the metal. It aligns the incentives. If the robot breaks, the vendor fixes it because they are still getting paid. It forces better reliability. Luna: That accountability is missing in traditional procurement. Once you buy it, it is your problem. Shifting that risk back to the provider changes everything. Lucas: It does. And it raises the question of what happens to the floor workers. If the robot is leased and owned by a third party, who is responsible when it malfunctions? The legal landscape is still catching up to this business model. Luna: Liability is always tricky when you introduce a third party into the physical workspace. I bet lawyers are having a field day with this. Lucas: Oh, absolutely. But let's pause here for a moment. If these conversations have moved your work forward in some small way, consider supporting the show directly. A handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that is literally what funds making this many of these. That is F. E. X. I. N. G. O. It keeps us ad-free and allows us to keep digging into these operational details without chasing clicks. Luna: Yeah, it really makes a difference for the quality of research we can do. Thank you to everyone who already supports us. Lucas: Back to the liability point. Because the robot is mobile and shares space with humans, the insurance premiums for these factories are actually rising in some regions. Insurers are unsure how to price the risk of a semi-autonomous system working alongside people. It adds another layer of friction to adoption. Luna: So you have integration costs, software subscriptions, and now higher insurance. The total cost of ownership is stacking up quickly. Lucas: It is. But the benefit is that you can scale up and down. If a contract ends, you return the robots. You are not stuck with a depreciating asset sitting in a warehouse for five years. That optionality has value that is hard to quantify but feels very real on the ground. Luna: Optionality is becoming the most valuable currency in manufacturing. Fixed assets feel risky when demand is this volatile. Lucas: Volatility is the key word. The factories that survive the next decade will be the ones that treat their production line like a software stack. Modular, replaceable, and adaptable. The cobot is just the tip of that spear. Luna: So the future isn't about building bigger factories. It is about building smarter, more fluid ones. Lucas: Fluidity wins. Whether it is cobots, AI scheduling, or modular supply chains, the ability to pivot without tearing down the walls is what separates the winners from the rest. Luna: Interesting. Do you think traditional heavy machinery makers will adapt to this lease model, or will new pure-play startups take over the market? Lucas: Both, probably. The incumbents have the engineering depth, but the startups have the agility. I suspect we will see a lot of acquisitions in the next two years as the big players try to buy the software layer. Luna: M&A activity in the automation sector. That is a trend worth watching. It changes the competitive landscape entirely. Lucas: It does. And it means that for factory managers, choosing a vendor today is a bet on their long-term viability. You don't want to build your entire operation around a startup that gets acquired and shuts down the platform. Luna: Vendor lock-in takes on a whole new meaning when the vendor might disappear. Due diligence is critical. Lucas: Absolutely. Check their balance sheet, check their roadmap, and make sure their software is open enough to switch if needed. Don't trust the demo. Trust the architecture. Luna: Practical advice. Always ask about the exit strategy before you sign the contract. Lucas: Especially when it involves millions of dollars of infrastructure. The exit strategy is often overlooked until it is too late. Luna: So, fluidity, optionality, and vendor stability. Three pillars for the modern factory. Lucas: Three pillars. And I think we will see more discussion on the data security side of these connected cobots next quarter. But that is a topic for another day. Luna: Data security in the physical world. Sounds like a complex rabbit hole. I am curious to see where that goes. Lucas: It is. But for now, keeping the floor moving smoothly is priority number one. Thanks for tuning in.