Autonomous Topography Detection and Traversal for an Inspection Robot Within the Beamline of Particle Accelerators

Particle accelerators feature ultra-high vacuum pipe systems with unique topography, i.e. with a multitude of different vacuum chambers of varying dimensions and varying pipe apertures. In order to be able to examine the interior of the entire vacuum system, even those parts which are not accessible without disassembling large parts of the accelerator, a semi-autonomous robot is being developed which shall traverse and visually inspect the vacuum system of particle accelerators. We present a generic concept based on distance sensors for the inspection robot to detect steps between vacuum chambers and gaps in the beamline. Movement strategies to autonomously overcome these basic obstacles are introduced. For evaluation we use simulations of ideal environments with flat surfaces as well as realistic beam pipe environments of the SIS100 particle accelerator. Additionally, a prototype of our robot concept confirms the implementation of all maneuvers. Results show that obstacles of previously unknown dimensions can be detected and reliably traversed.