Stanford researchers have developed Mobile-ALOHA, an open-source robotic hardware system capable of performing complex mobile manipulation tasks such as cooking, using elevators, and storing objects. The low-cost robot was trained using only 50 demonstrations and co-training of imitation learning algorithms with static ALOHA data, enabling it to autonomously perform complex tasks such as sautéing shrimp, storing heavy pots, and rinsing a used pan. The team behind Mobile-ALOHA, Zipeng Fu, Tony Z. Zhao, and Chelsea Finn, aims to make robotics more accessible and impactful by providing affordable hardware solutions. Mobile-ALOHA can carry 100kg and move up to 1.6m/s, while costing only $7,000, says Zhao. The team has open-sourced the hardware assembly tutorial, codebase, and all software and data related to Mobile-ALOHA.