May Mobility

May Mobility builds and operates autonomous shuttles designed to run in real-world conditions. Founded in 2017 by Edwin Olson, the company works with transit agencies and municipalities on public transportation services, and has delivered more than 300,000 autonomy-enabled rides across deployments in the United States and Japan, including Minnesota and Tokyo.

At the centre of its technology is Multi-Policy Decision Make (MPDM), a real-time AI system based on reinforcement learning. Rather than relying on a fixed rules-based stack, MPDM learns while driving and is built to handle scenarios that prescriptive systems cannot navigate. It generates new training examples every 200 milliseconds, a cadence that shapes the engineering problems the company works on in real-time AI, multi-policy decision making and self-driving systems.

Deployed autonomous vehicles sit at the intersection of several technical domains: autonomous vehicles, reinforcement learning, real-time AI and multi-policy decision making. The work is oriented towards deployable autonomy - technology intended for use in municipal transportation and urban mobility settings rather than for demonstration alone.

May Mobility describes its mission as making urban mobility safer, greener and more equitable. Its approach is partnership-based, working with transit agencies, municipalities and major platforms to integrate its shuttles into existing transportation networks.