On 13 November 2026, MIT News published an article about the company Robust AI, founded by MIT graduate Anthony Jules and professor emeritus Rodney Brooks. The company presented the mobile robot Carter — a wheeled cart designed for seamless interaction with people in dynamic warehouse and factory environments.
Carter is equipped with a force-sensitive handle, an AI-based computer vision system and holonomic wheels that allow it to move instantly in any direction. A person can take over control, push or pull the robot with one hand, or give a gesture or command via a touchscreen. This is fundamentally different from traditional industrial robots, which either work inside cells or require a full stop when a person appears.
The company's approach is built on the idea that success is determined not by the robot's autonomy in itself, but by the joint productivity of the "robot — human — backend" system. In real deployments, Carter assists with picking, moving loads and navigating workers, cutting a person's daily walking by an average of tens of thousands of steps. The robot can lead a person to the right cells or follow behind them during loading.
Robust AI's methodology emphasizes human-in-the-loop: the robot does not strive for full autonomy but leaves a high level of agency to the human. This reduces risks in unpredictable environments where layouts change and tasks require flexibility. Unlike approaches oriented toward maximum autonomy (for example, systems with deep reinforcement learning for managing robot traffic), the priority here is predictability and ease of human intervention.
In the robotics landscape, this contrasts with solutions where robots are either fully autonomous and require dedicated zones, or are heavy manipulators inside protective fencing. Carter occupies the niche of "collaborative mobility," where the design of the hardware (force feedback, holonomy) and software (vision and simple interfaces) is specifically tailored for working together without additional infrastructure.
For the industry, this means faster adoption: such robots can be deployed at existing sites without costly reconfiguration of space. This lowers the barrier to entry for mid-sized enterprises and changes the economics of automation — the focus shifts from replacing people to increasing their efficiency.
The question of long-term scalability remains open: how robustly the robot's behavior holds up under a sharp increase in the number of units and more complex scenarios. Independent research must show whether the level of safety and productivity is maintained under dense mixed traffic of people and robots in real-world conditions.
The main lesson of the development is that in complex physical environments, the winner is not the most autonomous robot, but the one that integrates best into the existing human system.
