Jacobi Robotics

Jacobi Robotics develops software for industrial robot arms, centered on a proprietary real-time motion planner that computes collision-free trajectories with computation times below one millisecond, removing the need for hand-coded waypoints and manual tuning. Its hardware-agnostic platform combines high-fidelity simulation and a digital twin, ultrafast motion planning, and reinforcement-learning agents so robots can make decisions within milliseconds and adapt to changing conditions. The company positions motion as the first of several building blocks in the Jacobi Platform, alongside planned components such as perception, grasping, and hardware-abstraction drivers.

The company's lead application is mixed-case palletizing. The Jacobi OmniPalletizer converts live, unsequenced mixed-case flow into store-ready pallets for retail, grocery, pharma, e-commerce, FMCG, and LTL operations, handling case flows from Pick-to-Belt, Goods-to-Person, or AS/RS systems and enforcing operational constraints including heavy-to-light stacking, crush limits, aisle-friendly grouping, and stop-based sequencing. Jacobi describes it as the first unsequenced mixed-case palletizing software deployed in production, and states that cells install into existing lines in brownfield facilities and can be operational within days. An earlier product, the Jacobi Palletizer, was announced in July 2024 as the first AI-powered solution built on the motion-planning technology.

Jacobi frames its market thesis around an automation gap: more than 2.5 million unfilled jobs across U.S. manufacturing, logistics, construction, and agriculture against roughly 40,000 robot arms deployed in 2022 and fewer than 35,000 new robots deployed in the United States in 2024. For its palletizing beachhead, it cites over 80 million cases stacked by hand daily in the U.S. and more than 80,000 reported injuries annually.

Founding story

The company was founded in Berkeley, California by roboticists and AI researchers from the Berkeley AI Research Lab, with the founding team drawn from UC Berkeley and Carnegie Mellon University. Forbes reports that Max Cao, Lars Berscheid, and Yahav Avigal met at the Berkeley AI Research Lab and started the company in 2022 based on their motion-planning research. The founding group also includes Jeff Ichnowski (Carnegie Mellon professor, former UC Berkeley postdoc) and Ken Goldberg (UC Berkeley professor and co-founder and chief scientist of Ambi Robotics).

Business model

Jacobi sells software rather than robot hardware, licensing a hardware-agnostic platform and productized applications such as the OmniPalletizer that run on third-party industrial robots and integrate into customers' existing lines and systems. Deployments feed a shared data lake that the company says improves its models across the fleet.

Traction

Robots running Jacobi software are described as deployed and performing tasks daily in active facilities, with the platform used by major system integrators and supporting multiple hardware OEMs. At its July 2024 launch the company reported agreements with US manufacturers, robotics solutions provider Formic, and Fortune 500 companies in electronics and CPG. Forbes reported customers including Formic plus electronics makers and consumer products companies, and total funding of $6.2 million. Subsequent milestones include a DHL innovation partnership awarded to a Fast Forward Challenge winner (January 2026) and 2026 collaborations with ABB Robotics and FORTNA.

Latest developments

In April 2026 Jacobi announced a collaboration under which ABB Robotics will integrate the OmniPalletizer AI software into its robotics portfolio and offer it to ABB's system integrator network as a productized path to mixed-case palletizing without upstream sequencing infrastructure, and a partnership through which FORTNA expands its automation portfolio into mixed-case palletizing using OmniPalletizer. In January 2026 DHL awarded an innovation partnership to a Fast Forward Challenge winner, and in February 2026 the company published material on why OmniPalletizer does not require perfect case sequencing.