Problem

Industrial-vehicle autonomy is still unusually expensive and slow to build. Robotics teams often have to recreate the entire stack—from path planning and controls to hazard detection and safety—for every tractor, forklift, yard truck, or mining vehicle. Polymath describes the problem as a lack of tooling that can turn an autonomy project into a reliable product, with development historically taking up to two years just to reach a demonstrable vehicle. The resulting economics are high engineering costs, long deployment cycles, and continued dependence on safety drivers or on-site operators.

The clearest use case is automating vehicles in controlled, closed environments where removing the driver has immediate operational value. Examples include yard trucks, mining trucks, agricultural equipment, logging vehicles, and airport ground-support equipment. In the ConGlobal application, Polymath’s collision-avoidance system was intended to meet safety requirements and enable safety-driver removal, improving the economics of remote or autonomous yard-truck operations.

Product / Service

Polymath provides a software-first autonomy and safety stack for off-highway vehicles. It combines AI and machine learning, controls, ROS, path planning, hazard and human detection, behavior trees, vehicle dynamics, and safety tooling, exposed through deployment infrastructure and a REST API. The platform is designed to be vehicle-agnostic and sensor-agnostic, while running autonomy on onboard compute to reduce networking and control-center costs.

The delivery model is a reusable platform that lets robotics teams focus on their industry-specific application rather than rebuilding basic autonomy. Polymath’s Caladan product provides simulated vehicles and environments that developers can control through an API, using Python rather than ROS, then transition the same behaviors to physical vehicles. The intended benefit is faster, more reliable deployment with fewer engineering resources and less need for safety drivers; the company says its software can support production-ready autonomy across many vehicle platforms.

Market

Polymath competes in autonomy software and safety systems for off-highway and industrial vehicles, serving equipment OEMs, fleet operators, logistics companies, and government or defense programs. The category includes both horizontal autonomy-stack providers and vertically integrated companies building autonomy for a particular machine or workflow. Relevant alternatives include specialized off-highway autonomy companies such as Built Robotics and SafeAI, as well as OEMs and customers’ internal engineering teams that build comparable systems themselves.

The available evidence indicates that Polymath is commercial rather than pre-revenue, although exact revenue, pricing, and customer counts are not disclosed. Its YC profile reported growing revenue and a growing robotic fleet, while the company has said customers were developing on its platform and that autonomous vehicles were operating unmanned for four to six hours per day. More recent company materials say the software runs across dozens of platforms and is trusted by top OEMs, large fleet operators, and government programs; its named ConGlobal collaboration and reported contract wins provide additional evidence of market traction.