Avea Robotics

Teams deploying robots need a way to control them remotely from day one, hand off to human operators when autonomy fails, and manage many robots across a fleet. Without shared infrastructure, companies must separately build teleoperation tooling, failure-recovery workflows, fleet monitoring dashboards, and pipelines to turn field incidents into training data. This likely means slower deployments, ad hoc tools per robot, and lost operational data when autonomy encounters edge cases it cannot resolve on its own.

Sentinel is a five-layer stack that connects a robot and its autonomy software to a human team: a robot interface layer (ROS 2/API) works with existing robots and software, a human control layer lets an operator see through the robot's cameras and drive it remotely over Wi-Fi or LTE, a hybrid autonomy layer hands control between autonomy and a human operator when the robot hits a failure it can't resolve, a fleet operations layer gives one view of every robot, operator, and alert, and a learning loop turns field interventions into episodes and datasets for retraining. Setup is a single installer with ROS 2, Python, and C++ interfaces, configured by pointing Sentinel at the robot's URDF, cameras, controls, and safety limits, then starting locally and enabling cloud and fleet tools as deployment scales.