Salem Robotics, a YC-backed startup, provides software that gives existing mobile robots the intelligence to perform complex inspection tasks in hazardous facilities. Founded by robotics researchers from UT Austin and Los Alamos National Laboratory, the team focuses on manipulation challenges like taking smears for nuclear contamination surveys or positioning detectors for leak detection.

The core problem they solve is enabling robots to execute precise, constrained movements—such as maintaining sensor orientation relative to a curved pipe—without requiring manually authored trajectories for every scenario. Their approach combines classical robotics (explicit geometry, planning, and control) with AI for semantic understanding of unfamiliar scenes.

The company believes that rather than building new hardware, they can add value by creating an application layer for existing platforms from companies like Boston Dynamics. This allows the same robot to perform diverse industrial procedures reliably, with theoretical guarantees where safety is critical.