LUS2 (Lumos Robotics) vs T800 (EngineAI)

LUS2 measures 160 x 50 x 30 cm and weighs 57 kg, making it more compact and lightweight compared to T800, which stands 173 cm tall and weighs 75–85 kg. LUS2’s smaller profile supports agility in dynamic environments, whereas T800’s larger magnesium-aluminum alloy exoskeleton is designed for durability and strength in heavy-duty tasks.

T800 is faster at 2 m/s or more versus LUS2’s 4 km/h (1.12 m/s), giving it quicker movement in indoor and complex environments. Both utilize visual SLAM and LiDAR mapping for navigation, but LUS2 also incorporates ultrasonic sensors for enhanced indoor localization. T800 emphasizes stability and continuous operation with an active cooling system in joints during demanding tasks.

LUS2’s sensor array includes Intel RealSense D435i, tactile sensors on fingers and palms, 360° LiDAR, ultrasonic sensors, IMU, gyroscope, force, and temperature sensors. T800 features RGB and stereo cameras, omnidirectional LiDAR, ultrasonic sensors, IMU, gyroscope, force, and temperature sensors. T800’s 360° omnidirectional perception system enhances continuous environment monitoring without blind spots.

Both robots support autonomous operation, learned behaviors, and teleoperation, integrating ROS2 and Python/C++ APIs. LUS2 provides proprietary OS with APIs tailored for large language model integration. T800 leverages advanced onboard AI computing with AGX Orin modules supporting 275 tera operations per second, enabling high-performance vision-language-action modeling, reflecting a heavier focus on complex AI tasks.

T800 has a solid-state lithium battery optimized for up to 4 hours of continuous operation per charge, supported by an active cooling system to maintain performance during intense use.

LUS2 is versatile across industrial automation, healthcare support, search-and-rescue, and research environments, benefiting from its lighter weight and faster speed. T800 is suited to heavy-duty industrial tasks, commercial retail, logistics automation, and human-robot interaction scenarios, with physical strength and endurance emphasized alongside specialized applications like robot boxing.

Both robots run proprietary operating systems while fully supporting ROS2 and programming in Python and C++. LUS2 includes APIs designed for integration with large language models, facilitating advanced conversational AI. T800 employs proprietary end-to-end vision-language-action AI models used by leading humanoid companies, emphasizing AI integration for manipulation and perception.

Safety measures in both robots include force limiting, collision detection, emergency stop capabilities, and collaborative modes intended for safe interaction in human environments, ensuring operational safety across their respective use cases.

Priced at $99,000, LUS2 is positioned as a cost-effective humanoid with a strong balance of specs suitable for varied applications. T800’s price of about $25,000 to $80,000 undercuts LUS2’s $99,000 despite its heavier build, advanced AI capabilities, and industrial focus, catering to users demanding higher strength and endurance.