On-demand urban transit vehicles with custom embedded Linux and real-time autonomy
Glydways builds autonomous on-demand transit vehicles designed for cities. The tech stack reveals a hardware-first company: custom Yocto Linux distributions, kernel drivers for Nvidia Tegra and NXP i.MX platforms, and real-time decision-making heuristics form the core. Active projects span board bring-up, simulation, fault handling, and dynamic vehicle allocation — all pointing to a company solving the embedded systems and operational complexity of deployed autonomous fleets at scale. The hiring mix (17 engineers, mostly senior) against 51–200 total headcount signals a deep engineering org building and iterating on physical systems.
Glydways designs high-capacity, on-demand urban mobility vehicles positioned as affordable to operate and profitable to run. Founded in 2016 and based in South San Francisco, the company operates as a privately held manufacturer and systems integrator. Revenue comes from city contracts and fleet deployments. The product stack includes custom embedded Linux, proprietary dispatch algorithms for real-time passenger allocation, and onboard fault detection and recovery. Current pain points center on safety regulation alignment, securing and scaling city contracts, and operational reliability — preventing double assignments and maintaining a single authoritative dispatch system across deployed fleets.
C, C++, Python, Ruby, Node.js, and .NET across embedded systems (kernel drivers), backend services, and frontend tools. Core vehicle logic runs on custom Yocto-based Linux on Nvidia Tegra and NXP i.MX processors.
Custom Linux distribution maintenance, kernel driver development for vehicle compute platforms, simulation and metrics, dynamic vehicle allocation algorithms, onboard fault handling, and real-time decision-making heuristics for dispatch optimization.
Other companies in the same industry, closest in size