We've joined NVIDIA Inception.

The hard part of offline autonomy was never the AI. It’s running that AI on something small enough to carry, frugal enough to run on a battery, and stubborn enough to keep thinking when the network is gone.

This week UBR was accepted into NVIDIA Inception — NVIDIA’s programme for deep-tech startups. For us, it’s about the tools that put capable models on low-power silicon: the Jetson edge modules our field brain is built to run on, and the toolchain — CUDA, TensorRT — that makes real-time inference work at the edge.

It changes nothing about the thesis and plenty about the pace. The field brain still senses, decides and acts entirely on-board, with no cloud and no link to lean on. The programme just helps us get there on better hardware, sooner — and test more before any of it reaches the field.

Same direction. Sharper tools.

If you’re building autonomy for the places the network can’t reach, talk to us.