Victor Boyd built an autonomous forklift from scratch — and deployed it in a pharmaceutical warehouse

Cavalla Industries

Victor Boyd dropped out of the University of Alabama at Birmingham and moved to San Francisco to build autonomous forklifts. His company, Cavalla Industries, is already deployed in a pharmaceutical warehouse — one of the most unforgiving environments in industrial logistics, where a single damaged pallet can run into hundreds of thousands of dollars.

Why build a new forklift from scratch

Cavalla's starting point was a retrofit approach, the same bet most competitors have made. It failed twice. First as a universal kit, then as a single-model retrofit. The problem is that forklifts, unlike cars, were never standardized around a common protocol. Two forklifts of the same model and year can have meaningfully different internals, making software-only solutions unreliable and too slow for customer requirements.

Boyd's answer was to build the forklift itself, working through a manufacturing partner in Q4 of last year. Off-the-shelf components cover only the motor. Everything else is Cavalla's own design, which Boyd argues is the only way to maintain full control, iterate quickly, and avoid the trade-offs baked into third-party hardware decisions.

“We realized that, you know, that was gonna be the thing for autonomous forklifts. You wanted to make it viable for customers, you actually had to make the platform viable for autonomy first. So we had to build our own forklift. And we did. We're deployed in a very difficult environment. We're doing better than anybody else in terms of throughput and reliability.”

Pharma as a proving ground

The pharmaceutical deployment is deliberate. Boyd treats the high-stakes environment as a forcing function — a damaged pallet in a standard consumer-goods warehouse might cost a few thousand dollars, but pharmaceutical product loss runs far higher. Surviving there, without causing damage, is the credential he plans to show every other potential customer who has been burned by autonomy vendors before.

By his own description, Cavalla is outperforming competitors on throughput and reliability at this site. The 2025 focus is making the product scalable rather than expanding deployments.

Teleoperation

About 50% of current operations run under teleoperation, using a literal Xbox controller operated by remote staff thousands of miles away. Boyd is unapologetic about it. Customers care whether the work gets done, not whether the system meets some internal autonomy threshold. The goal is to push autonomy higher over time to improve margins, but the customer metric is task completion.

The longer ambition

Right now, operators assign the forklift a fixed daily task — moving specific pallets between designated zones — and the system executes without further instruction. Boyd wants to move toward running the entire warehouse physical operation autonomously, including inventory management, staging for incoming trucks, and proactive decision-making without customer input. The forklift is the wedge into a broader warehousing operating system.