Welcome, Asimov 1!
Asimov 1 is an open source educational humanoid from MenloAI that brings real robot hardware into the hands of developers, researchers, and robotics students. Standing 120 cm tall and weighing 35 kg, it offers 27 total degrees of freedom in a compact full body format, with a top speed of 3 km per hour, a strength figure of 3 kg, and about 2 hours of runtime per charge. The system is sold as a build it yourself platform priced at 20,000 USD, turning assembly into part of the learning experience rather than hiding the machine behind sealed enclosures. Its current Humanoid Guide score of 2 places it in the early stage tier, with a Navigation rating of 1 and a Manipulation rating of 1, so it is best understood as a practical research platform for foundational mobility and control work. It is safe with humans, runs Asimov OS on Raspberry Pi 5 and Radxa CM5 hardware, and supports LLM integration for experiments that connect physical action with higher level intelligence.
Asimov 1 is strongest as a hands on platform for testing basic autonomy in real environments. Its present capability set includes obstacle avoidance, which makes it useful for studying perception, route selection, and low speed navigation around objects in labs, classrooms, and structured indoor spaces. It also supports sorting goods, giving teams an accessible way to prototype simple logistics and handling workflows while tuning policies, sensor processing, and task sequencing. A 2 MP monocular camera supplies visual input, while six internal CAN bus branches, Bluetooth, Ethernet, and WiFi provide flexible integration with external tools and data pipelines. With LLM integration available, users can connect language driven interfaces to concrete robot behavior, creating educational demonstrations and early research systems that bridge instruction, sensing, and action.
Asimov 1 is already in production and represents a distinctive contribution from Singapore to the global humanoid field. Its open source hardware and software model gives universities, independent builders, and robotics teams a transparent alternative to closed commercial systems, especially for courses and labs that value direct access to mechanics, electronics, and control stacks. MenloAI extends that accessibility with Digital Asimov, a browser based simulator that helps users explore the design and test ideas before working on the physical robot. That combination of real hardware, learn by building assembly, and openly inspectable software makes Asimov 1 significant beyond its current skill score. It serves as a training ground for future humanoid engineers and a practical entry point for organizations building internal expertise in embodied AI.
