New App Lets Anyone Operate a Robot From Their Phone

Why It Matters

COBALT lowers the barrier to robot teleoperation, accelerating data‑driven AI training and opening new remote‑work and education markets for the robotics industry.

Key Takeaways

• COBALT lets smartphones control robot arms via WebRTC.
• Study involved users from nine countries operating remotely.
• Data collected enables training of autonomous robot policies.
• Platform could power a gig‑economy “Uber for robots.”
• Phones preferred over VR or controllers for teleoperation.

Pulse Analysis

The COBALT platform leverages the ubiquity of smartphones to solve a long‑standing robotics challenge: intuitive, low‑latency teleoperation. By streaming video and control signals over Web Real‑Time Communication, the system delivers sub‑second feedback, a critical factor for precise manipulation. Unlike VR headsets or specialized controllers, a phone’s built‑in sensors provide a natural, familiar interface, reducing training time and expanding the pool of potential operators. This technical simplicity positions COBALT as a scalable bridge between human intuition and robotic precision.

Beyond the hardware breakthrough, COBALT’s real value lies in its ability to harvest massive, diverse datasets for policy learning. The research team collected operation logs from participants across Indonesia, India, Pakistan and other nations, demonstrating that non‑expert inputs can match the quality of traditional lab‑generated data. Such crowdsourced data pipelines address the bottleneck in robot‑centric AI, where simulated environments fall short of real‑world variability. By democratizing data collection, the technology could accelerate the rollout of autonomous manufacturing lines and service robots, reshaping supply‑chain dynamics.

The commercial implications are equally compelling. A smartphone‑based teleoperation marketplace could emerge, allowing individuals to earn by remotely guiding assistive robots in homes or factories—essentially an “Uber for robots.” Companies could tap this on‑demand labor to handle edge cases that automated systems struggle with, improving safety and uptime. Educational institutions can also integrate COBALT into curricula, giving students hands‑on experience without costly hardware. As the platform matures, it may redefine both the gig economy and the economics of robot deployment, driving broader adoption across sectors.