The robot was designed by a team at Carnegie Mellon University (CMU). They used a simulation to teach it how to perform its task. The team used an AI-driven approach called ‘reinforcement learning’. The robot was rewarded each time it correctly placed a shirt further along an arm.