RNoid

The RNoid (also styled as R Noid or R-Noid) from Robot.com represents a practical, deployment-ready humanoid robot optimized for service-oriented tasks in indoor environments such as hospitality, reception areas, workplaces, light logistics, and educational settings. Launched as part of Robot.com's broader robotics ecosystem in late 2025—following a rebrand from Kiwibot—the RNoid stands out for its emphasis on real-world utility over spectacle, with over 500 units in the company's fleet completing more than 1.7 million tasks across the US, Canada, UAE, and Saudi Arabia by year-end. Its architecture centers on a lightweight 50 kg aluminum alloy frame with ABS composite shells, achieving IP32 ingress protection for dust and light water resistance. At 160 cm tall, it navigates via smooth omni-wheels reaching 6 km/h, earning a navigation performance score of 3/5, while manipulation scores 2/5 on the H.G scale. Core hardware includes 22 degrees of freedom overall, with dual 7 DOF arms (14 DOF total for arms) delivering up to 5 kg (11 lbs) payload per arm or ~9 kg dual-arm, enhanced by 10 DOF dexterous 5-fingered hands. Propulsion uses brushless motors with harmonic drives in arms for precision and planetary gears in joints. Vision relies on 1080p cameras, supported by advanced sensors for human-safe operation. Compute is ARM-based CPU paired with an embedded AI accelerator, running a Linux base layered with Robot.com's Remi OS—a customized platform integrating L4 autonomy, tele-ops, and fleet data management for real-time optimization. Connectivity spans Wi-Fi, Bluetooth, and 5G, enabling low-latency (250-450 ms glass-to-action) cloud-based LLM integration for intuitive dialogue and task comprehension. AI architecture leverages cloud LLMs for natural communication, with onboard processing handling navigation, obstacle avoidance, and manipulation planning. The dual expressive screens facilitate branding, emotional cues, and user interaction, while modular end-effectors (gripper or Dex-Hand) allow task adaptation. Thermal management and 6-hour runtime per charge support extended shifts, with self-charging options in fleet deployments. Real-world deployments highlight RNoid's maturity: in GXO warehouses, it autonomously handles semiconductors, boxes, and parts alongside humans; in kitchens/service stations, it performs repetitive precision work; high-traffic areas see it for inspections and digital ads; campuses utilize it for L4 food/package delivery. Trusted by Fortune 500s like Sodexo, Grubhub, Aramark, AWS, and semiconductor firms, it exemplifies scalable, ROI-positive automation. Unlike research prototypes, RNoid's production-ready status, verified through 2025 milestones, positions it for hospitality front-desks, workplace assistance, and light logistics, fostering human-robot collaboration without replacement. Future scalability draws from fleet learning, where remote supervision (5-10 robots per operator) ensures reliability in dynamic environments. (612 words)