The paper introduces Mobile ALOHA, a low-cost, bimanual, whole-body teleoperation system for mobile manipulation tasks. The system, costing $32k, includes a mobile base and two arms, allowing for independent movement of the base while both arms can be controlled simultaneously. The authors present a co-training approach that leverages existing static ALOHA datasets to improve the performance of Mobile ALOHA on complex mobile manipulation tasks. By training with 50 demonstrations per task, Mobile ALOHA can achieve success rates of up to 90% on tasks such as sauteing shrimp, opening cabinets, calling elevators, and rinsing pans. The paper also discusses the hardware design, co-training methodology, and experimental results, demonstrating the effectiveness of Mobile ALOHA in performing a wide range of household tasks.The paper introduces Mobile ALOHA, a low-cost, bimanual, whole-body teleoperation system for mobile manipulation tasks. The system, costing $32k, includes a mobile base and two arms, allowing for independent movement of the base while both arms can be controlled simultaneously. The authors present a co-training approach that leverages existing static ALOHA datasets to improve the performance of Mobile ALOHA on complex mobile manipulation tasks. By training with 50 demonstrations per task, Mobile ALOHA can achieve success rates of up to 90% on tasks such as sauteing shrimp, opening cabinets, calling elevators, and rinsing pans. The paper also discusses the hardware design, co-training methodology, and experimental results, demonstrating the effectiveness of Mobile ALOHA in performing a wide range of household tasks.