Researchers at the Massachusetts Institute of Technology have developed a system-on-a-chip that enables tiny autonomous robots to generate detailed three-dimensional maps of their surroundings while consuming only a fraction of the power required by existing technologies. The innovation addresses one of the biggest challenges facing small robots and battery-powered devices: performing sophisticated navigation without the energy demands of conventional mapping systems.
Small robots designed for applications such as industrial inspections, infrastructure monitoring, or search-and-rescue missions often operate under strict power and memory constraints. Traditional simultaneous localization and mapping (SLAM) techniques typically rely on dense voxel-based representations of the environment, which require substantial memory and computational resources. The MIT team tackled this limitation through a hardware-software co-design approach, developing both a highly efficient mapping algorithm and dedicated hardware optimized specifically for that workload.
Instead of representing environments with conventional voxel grids, the chip uses compact Gaussian ellipsoids to model surrounding objects. These adaptable geometric representations capture curved surfaces more efficiently while dramatically reducing memory requirements. The specialized hardware accelerates the algorithm directly on the chip, allowing robots to construct accurate 3D maps in real time while consuming only about six milliwatts of power, roughly equivalent to the energy used by a single LED.
The low-power design expands the technology’s potential beyond robotics. Battery-constrained devices such as lightweight augmented reality headsets could use the chip to create real-time spatial maps for applications including medical training, equipment maintenance, and industrial assembly without sacrificing battery life. By minimizing energy consumption while maintaining mapping accuracy, the technology supports longer operating times and more compact device designs.
The research demonstrates the value of jointly designing algorithms and hardware rather than optimizing them independently. According to the researchers, this integrated strategy enables the system to store large environmental maps in very little memory while maintaining exceptional energy efficiency. As miniature robots become increasingly important for inspecting hazardous spaces, navigating confined industrial environments, and supporting autonomous operations, the new chip could provide a practical foundation for intelligent navigation on devices where power, size, and weight are critical engineering constraints.