He built a robot to deliver your packages. The world wanted the box handler instead. So he listened.
Walk into any parcel depot and you will find a conveyor belt doing what conveyor belts have always done: moving flat, in a line, one box behind another. Austin Oehlerking looked at that belt and asked a stranger question. What if it could go up?
Boxbot, the company he co-founded in 2016 and runs as CEO out of Alameda, California, builds an AI-driven storage machine that stacks, sorts, and sequences packages in three dimensions instead of two. In his own framing, the pitch is almost rude in its simplicity: take the flattest, dumbest object in any warehouse and turn it into something three-dimensional. The result is a modular system that stores size-agnostic, high-throughput payloads in a fraction of the floor space, then hands each parcel back exactly when it is needed.
The same base module drops into three very different rooms. A last-mile delivery center, where trucks need loading in a precise sequence. A bonded customs warehouse, where goods sit under regulatory lock. A pickup locker, where a shopper wants one specific box, right now. One machine, three businesses - which is the kind of leverage a hardware founder dreams about and rarely gets.
None of this was the original plan. And that is the interesting part.
The first scalable application of self driving technology will be the delivery of things, not people.
Boxbot started life as a stealthy self-driving startup. The bet, laid out above, was that autonomous vehicles would prove themselves first on cargo, not commuters - that a robot van creeping through Oakland with a load of parcels was a more tractable problem than a robotaxi. Toyota AI Ventures put in $7.5 million. Pear Ventures handed the team $250,000 after they won the 2017 UC Berkeley Startup Challenge. The self-driving delivery future looked funded and close.
Then customers did something founders never quite plan for: they got excited about the wrong thing. The van, the whole autonomous centerpiece, drew polite nods. The box-handling robot bolted to it drew questions, meetings, and follow-ups. Oehlerking noticed that while everyone loved the long-term romance of autonomy, the actual near-term dollars were pooling somewhere less glamorous - warehouse automation. So in 2023 he pivoted the company toward the machine people actually wanted to buy.
There was plenty of enthusiasm for the long-term potential of autonomous technology, but a lot of the near-term capital investment seemed to be concentrated within warehouse automation tech.
The reward for listening arrived as a $12 million Series A led by Playground Global, with Maersk Growth, Toyota Ventures, Pear Ventures, and Artiman Ventures alongside. Total funding climbed to roughly $28 to $29.5 million. Richard Peretz, the former CFO of UPS, took a board seat - a legacy-logistics heavyweight signing on to a 20-person robot company, which tells you something about who is paying attention.
Oehlerking's engineering runs deep and slightly obsessive. He earned a mechanical engineering degree from MIT in 2008, then went back for a master's in the same field, steeping himself in power electronics and microcontroller programming. Later came an MBA from Harvard Business School. Along the way he built robots and electric vehicles at MIT's Field Intelligence Lab, ran vehicle engineering at the electric-commuter startup Arcimoto, and spent 2011 to 2014 at Tesla as a senior software engineer and team lead, working on the RAV4 EV and the Model S.
His first company was less dignified. In 2005 he co-founded SharpEdge Studios, a casual video game with a virtual economy and microtransactions - a concept a few years ahead of the mobile platforms that would have made it work. It also had five co-founders, which turned into exactly the equity mess you would expect. The lesson stuck: never split a company that many ways again. His Boxbot partnership is a tidy two-hander with Mark Godwin, a Berkeley PhD who built drones for the Office of Naval Research and raced in the DARPA Urban Grand Challenge.
The drive traces back further, to a household where nobody had a college degree and his father sweated quarterly sales targets in enterprise software. Watching that stress, Oehlerking decided he would rather own the business than carry someone else's number. It took him to MIT - twice - and then to a robot company that is quietly rebuilding the plumbing of e-commerce.
Boxbot's insight is that most warehouse space is wasted on air above the belt. Give the parcels a third dimension and the same footprint holds far more, retrieved on demand.
Size-agnostic parcels load into a dense, modular gantry - vertical space that a flat conveyor never used.
An AI operating system decides what comes out when, buffering volume and smoothing labor across the day.
Each box is handed back on demand - to a truck, a customs officer, or a shopper at a pickup locker.
Interns at Raytheon; founds MIT's ThinkBIG Club to gather student entrepreneurs when startups were still uncool.
Co-founds SharpEdge Studios - a microtransaction video game, five founders, and a masterclass in what not to do.
Graduates MIT with a mechanical engineering degree; later returns for a master's in power electronics.
Director of Vehicle Engineering at Arcimoto, the Oregon electric-commuter-vehicle startup.
Senior Software Engineer and Team Lead at Tesla, shipping software for the RAV4 EV and Model S.
Co-founds Boxbot with Mark Godwin as an autonomous last-mile delivery company.
Wins Pear Ventures' UC Berkeley Startup Challenge and a $250,000 prize.
Pivots Boxbot to AI-driven 3D parcel storage; closes a $12M Series A led by Playground Global.
The first scalable application of self driving technology will be the delivery of things, not people.
There was plenty of enthusiasm for the long-term potential of autonomy, but the near-term capital was concentrated in warehouse automation.