Latest / The Robotics Podcast with Fexingo: Autonomous Systems, Industrial Robots, and Hardware / The Cost of Cobots in Small Factories
Transcript
- Lucas: There is a specific kind of optimism that has been bleeding out of the industrial robotics sector lately, and it isn't about the technology itself. It is about the math. For the last five years, every vendor pitch I have sat through has promised that collaborative robots, or cobots, would pay for themselves in under two years. That was the golden rule: buy the arm, install it, watch your margins expand. But if you look at the actual deployment data from this quarter, that timeline is stretching dangerously long. Luna: I have seen some reports suggesting that integration costs are eating up the savings, but is it really that bad for smaller operations? I thought cobots were supposed to be the democratization of automation. Lucas: They were supposed to be. The whole value proposition of a company like Universal Robots or Techman Robot was that you could plug them in without needing a team of electrical engineers. But here is the thing that nobody wants to put in the brochure: the robot is only twenty percent of the cost. The other eighty percent is the cage-free safety sensors, the end of arm tooling, and the hours of programming required to make it behave predictably in a messy human environment. Luna: So you are saying the hardware is cheap, but making it work is expensive? That sounds like a classic case of hidden operational overhead. Lucas: Exactly. Let me give you a concrete example from a packaging facility we audited last month. They bought two six-axis cobot arms to handle palletizing. The sticker price for the arms and basic controllers was roughly forty thousand dollars each. Very reasonable. But to get those arms to work safely alongside human packers without light curtains and fencing, they had to install LiDAR scanners, force-torque sensors, and custom grippers. By the time the systems integrator finished the job, the total spend per arm was closer to one hundred and twenty thousand dollars. Luna: That is a tripling of the budget. Did the output justify that kind of spend, or did the return on investment just not materialize? Lucas: The output was fine. In fact, it was better. These two arms increased their packing speed by thirty-five percent and reduced product damage by half. The problem was that the labor savings didn't come fast enough to cover that initial one hundred and twenty thousand dollar hit. Normally, you would recoup that in eighteen months. Because of the integration bloat, their break-even point pushed out to three years and four months. For a small business with tight cash flow, that is a lifetime. Luna: Three years is a huge risk when interest rates are where they are now. Are these companies regretting the purchase already, or are they just stuck waiting for the payoff? Lucas: They are stuck. And what is interesting is that the maintenance burden is also higher than anyone anticipated. A traditional industrial robot in a caged cell runs for ten years with minimal intervention. A cobot working next to humans takes a beating. Sensors get knocked, grippers wear out faster because they are handling irregular consumer goods instead of uniform automotive parts, and the software needs constant tweaking as the production line changes. We are seeing service contracts cost nearly fifteen percent of the original asset value annually. Luna: It feels like the market is realizing that flexibility comes with a premium. You can't just set it and forget it anymore. Lucas: Right. The old model was rigid but cheap to run. The new model is flexible but expensive to maintain. And for small manufacturers, who don't have in-house robotics teams, that maintenance becomes a bottleneck. They hire a general mechanic, train him to fix the cobot, and suddenly their labor costs go up again because that mechanic is spending forty hours a week debugging a gripper instead of running the press. Luna: So the very people who need automation the most—the ones short-staffed and struggling to find reliable workers—are the ones least equipped to manage the complexity of these systems. Lucas: That is the irony. The labor shortage drove the adoption of cobots in the first place. Companies couldn't find people to do repetitive, dull tasks. So they bought the robots. But now they are finding that managing the robots requires a different kind of skilled labor that is equally hard to find. It is not a blue-collar shortage anymore; it is a brown-collar shortage of technicians who understand both mechanics and code. Luna: I wonder if the vendors are going to pivot their messaging. If the hardware is no longer the selling point, what are they pushing? Lucas: They are pivoting hard into managed services. You are starting to see companies offer the robot installation for free, but then lock you into a five-year software and maintenance subscription. It shifts the risk from the buyer to the seller, which is smart for the vendor, but it means the customer never truly owns the asset. It turns a capital expenditure into a recurring operating expense, which changes how CFOs view the deal entirely. Luna: That sounds like the software as a service model invading the factory floor. Is that sustainable for a business that might go under before the contract ends? Lucas: It is risky. If the factory closes, you lose the equipment, but you still owe the subscription fees depending on the contract structure. We saw one case where a small injection molding plant folded after two years, but they were still billed for the cloud-based monitoring suite attached to their cobots. It left them with a liability that had no tangible asset backing it. It is a financial trap for unprepared founders. Luna: Wow. So the lesson here is that you cannot just buy a robot and hope for the best. You need a whole ecosystem around it. Lucas: Precisely. And that ecosystem is currently fragmented. You have the robot maker, the integrator, the sensor supplier, and the software platform, and none of them talk to each other efficiently. The customer is left holding the bag when things break. We are looking at a correction phase where many small factories will try to sell these cobots on the secondary market, likely at a steep discount, because they simply cannot justify keeping them online. Luna: Do you think the big players are going to step in and buy these assets cheaply, consolidating the market? Lucas: Maybe. Or maybe they will just ignore them. Large manufacturers have their own internal engineering teams and standardized processes. They don't need third-party cobots. They build their own automation lines. The cobot market is really only viable for the mid-market segment, and that segment is currently getting squeezed between rising input costs and stagnant demand for consumer goods. Luna: It seems like the hype cycle peaked too early. Everyone wanted to be an automated factory overnight, but the infrastructure wasn't ready. Lucas: I agree. The technology matured faster than the business models did. We assumed that once the robot could pick and place, the economic case was closed. We forgot about the context. The robot doesn't operate in a vacuum. It operates in a facility with power constraints, space constraints, and human dynamics. Ignoring those variables leads to exactly the kind of cost overruns we are seeing now. Luna: So for listeners thinking about bringing automation into their shops, what should they be doing differently today compared to two years ago? Lucas: Stop looking at the robot price tag. Look at the total cost of ownership. Ask your integrator for a detailed breakdown of the safety systems, the training hours, and the expected downtime. Get those numbers in writing. And honestly, consider if you really need full automation. Sometimes, a semi-automated solution where the human does the complex part and the machine does the heavy lifting is more resilient and cheaper to maintain. Luna: Hybrid workflows. That makes sense. Less dependency on perfect machine performance, more reliance on human adaptability. Lucas: Exactly. Humans are terrible at repeating the same motion ten thousand times, but they are amazing at adapting to a weird box shape or a loose cable. Cobots are great at the repetition. Use them for what they are good at, and don't expect them to solve your entire labor crisis. Luna: It is a reminder that tech is a tool, not a magic wand. If you don't understand the mechanics of the tool, you will pay for it twice. Lucas: That is a fair way to put it. You pay for the hardware once, and then you pay for the ignorance of how it works repeatedly until you figure it out. Luna: Before we move on to the next topic, I want to pause for a second. This conversation about the real costs of tech adoption is something we dive deep into because it matters for your bottom line. Honestly, if today was worth a coffee to you, that is the link — buy me a coffee dot com slash fexingo. It keeps this network ad-free and allows us to keep digging into the unsexy details of business. Lucas: Yeah, support is what lets us keep these episodes coming. Thanks for listening, Luna. Let's get back to the supply chain implications. Luna: Right, so if small factories are scaling back on these robotic investments, how does that ripple out to their suppliers? Lucas: It creates a bottleneck at the lower end of the manufacturing spectrum. If small component makers can't automate, they become slower and more expensive. That pushes orders toward larger, fully automated plants. We are seeing a consolidation trend where mid-sized jobs are being swallowed by big players because the big players have the balance sheet to absorb the integration costs that the small guys can't. Luna: So the barrier to entry for manufacturing is actually rising, even though the robots are supposed to make it easier? Lucas: It is ironic, but true. The dream was that automation would let anyone start a factory. The reality is that it is becoming a moat for those who can afford the complexity. Until the software gets simpler and the hardware gets cheaper, we are going to see fewer, but much larger, manufacturing hubs. Luna: It leaves me wondering if there is a middle ground. Maybe open-source robotics platforms will emerge to lower that technical barrier? Lucas: Possibly. There is some movement in open-hardware initiatives, but proprietary ecosystems are sticky. Once you are locked into a specific vendor's software language, switching costs are high. We will have to wait and see if the market demands standardization, or if fragmentation remains the norm. Luna: Whatever happens, the era of easy automation is definitely over. It is going to require more brains and less brute force. Lucas: Well said. And for any founder listening, that is a challenge, but also an opportunity. Those who master the complexity will win. Those who just buy the shiny object will struggle. Luna: Let's see if the next quarter brings any new tools that actually simplify this process, or if we are stuck in the weeds a bit longer. Lucas: I doubt it will be quick. Complexity rarely resolves itself overnight. But we will keep tracking it.