As winter approaches, here's a story about why hardware is hard. ❄️🥶 About a year ago, we started getting reports from the field about undesirable behavior when our robots were turned on. They would behave unpredictably.

My biggest takeaway from #icra2025 was just how many students want to found a robotics company. But building a company is much more than building a robot. Instead of focusing on the HOW of robots, I encouraged people to think about the WHY. What problem do you solve, and for who?

I'm 100% in favor of purpose-built robots. While AI could enable humanoids that do lots of things imperfectly, specialized form factors are the only path to a sustainable robotics business. Thanks to everyone who turned out to see Dusty at the Cerebral Valley AI Summit.

Replit left San Francisco for Foster City. The "why" we're leaving is boring, sad, and predictable (crime, dysfunction, etc), so instead let me tell you why we chose Foster City. Foster City embodies the American post-war optimism and the long-lost California pro-growth

Love this story and want to be a part of this? We're hiring.

Convinced that Dusty would be an integral part of all their upcoming warehouse build-outs, they inked a 7-figure enterprise contract for multiple robots deployed over the next three years. This is how we win.

The client was sold. With our accuracy, they gained a certainty they had never had before in the correctness of the model and how that translated out to the field. Rather than wasting time figuring out who made a mistake, they had certainty that everything was in the right place.

Once everyone calmed down, it was clear that the model was at fault, not the robot. The project team wouldn’t have discovered this bust for weeks using manual layout. Andrew had us re-mobilize on the next warehouse build. The second trial, and the third, went perfectly.

The subcontractor had followed the model, but the model was wrong. When the concrete pad location was overlaid on top of the anchor bolt locations, sure enough they overlapped -- which is exactly what Dusty showed in the field.

Our CS team quickly convened online to diagnose. Robots don’t make mistakes. Most likely there was an error in the 3D model that caused this problem. Sure enough, the location of that concrete pad was incorrectly specified in the 3D model.

It looked like Dusty’s robot had made a mistake. What’s more, this mistake cast into question everything that was printed. The field crews started cursing out poor Nate, in both English and Spanish. He was asked to leave the site and not return.

Until: a problem. Anchor bolt locations were printed on top of a reinforced concrete pad that would be used to support heavy machinery. The facility had been carefully designed -- and no anchor bolts were supposed to be on those pads. “Shut it down!” the superintendent yelled.

The site team was amazed. “This is the future of layout for sure,” Andrew said. Everyone was patting each other on the back, excited to be part of this momentous occasion.

The project team was excited to see Dusty in action. Andrew the PM welcomed Nate to the site and introduced him to the site superintendent and the operator who would be trained on FieldPrinter. Training went quickly and we started printing points at Dusty’s usual rapid clip.

Our VDC team worked with the client to prepare robot-ready drawings. This was one of the more complex jobs we’d been involved with, due to the sheer amount of layout in such a dense facility. Finally our CSM Nate brought a FieldPrinter to the site in Mississippi.

I almost didn't take the call. Robotic warehouse automation was not in our ICP. Luckily my team overrode me and took the meeting. The automation company was intrigued, and we agreed to an annual contract with a pilot period to try a FieldPrinter on their next build-out.

Cue Dusty. Their project manager had heard of a robotic solution for layout, and wanted to know more. @DustyRobotics

Traditional layout is labor intensive and takes months. They didn't have months. As with any manual process, occasional errors increase delays and cost. Each day these warehouses are not in operation puts these retailers’ revenue at risk.

It was the fall of 2022. A robotics company had reached out to us with a problem: they install robotic automation in distribution centers for global retailers, and they needed to mark out where miles of machinery and product shelving were to be installed with mm precision.

First, we got ourselves kicked off the construction site. Then, they signed a 7-figure enterprise contract. Here's how we did it. ↓

Venture must be coming back. I've lost count of the number of new founders reaching out to me for advice on company formation and fundraising.