Probabilistic robotics is at home in the fields of Artificial Intelligence, the Internet of Things, and robotics. It describes methods by which robots can perceive their environment and navigate within it, despite uncertain or incomplete information. This means that robots do not make fixed assumptions, but rather consider that sensors sometimes provide inaccurate data and that decisions must be made under uncertainty.
A simple example: a robotic vacuum cleaner is tasked with cleaning a flat. Its sensors detect pieces of furniture, but sometimes objects are misidentified or overlooked. Using probabilistic methods, the robot can learn from experience how likely a particular situation is and react flexibly, rather than rigidly adhering to a predetermined plan. This allows it to better avoid obstacles or reorient itself if it gets lost.
Probabilistic robotics plays a key role in the development of modern, intelligent robots that operate reliably and safely in changing environments – such as in factories, logistics, or the smart home. The technology makes robots more adaptable and robust against errors and uncertainties.
Deutsch 
English (UK)