Can We Learn From Insect Navigation to Make Better Autonomous Robots?
TU Delft researchers have drawn inspiration from ants’ navigation techniques to develop an autonomous navigation strategy for tiny, lightweight robots that could revolutionise industries and research environments.
Navigating complex lab and research spaces has always presented significant challenges, but the innovative approach inspired by insect navigation may provide an effective solution. Recent findings by TU Delft researchers, published in Science Robotics, reveal an insect-inspired strategy that allows tiny robots to autonomously return home after long trajectories with minimal computation and memory requirements.
Tiny robots, ranging from tens to a few hundred grams, have immense potential for real-world applications. Their light weight ensures safety even in accidental collisions, and their small size enables navigation in narrow areas. Cost-effective production could allow large-scale deployment, covering expansive areas swiftly, such as in greenhouses for early pest or disease detection.
The primary challenge for tiny robots is autonomous operation due to their limited resources compared to larger robots. Traditionally, autonomous navigation for small robots relied on external infrastructure like GPS or indoor wireless beacons. However, GPS is often unavailable indoors, and beacons can be costly and impractical in certain scenarios, such as search-and-rescue missions.
Large robots, like self-driving cars, use AI for autonomous navigation, relying on heavy sensors like LiDAR, which tiny robots cannot carry or power. Vision-based approaches, although power-efficient, typically create detailed 3D maps of the environment, demanding significant processing and memory, unsuitable for tiny robots.
This is where nature provides valuable insights. Insects operate over distances relevant to many real-world applications using minimal sensing and computing resources. They combine odometry, tracking their own motion, with visual navigation strategies. Ants, for instance, take occasional “snapshots” of their environment and navigate by comparing these snapshots to their current surroundings, similar to Hansel and Gretel’s breadcrumb trail in the fairy tale.
“Snapshot-based navigation can be compared to how Hansel tried not to get lost in the fairy tale of Hansel and Gretel. When Hans threw stones on the ground, he could get back home. However, when he threw bread crumbs that were eaten by the birds, Hans and Gretel got lost. In our case, the stones are the snapshots,” explained Tom van Dijk, first author of the study.
The TU Delft team’s innovative strategy spaces snapshots further apart, relying on odometry between them. This approach allows a 56-gram “CrazyFlie” drone, equipped with an omnidirectional camera, to cover distances of up to 100 metres with only 0.65 kilobytes of memory. All visual processing occurs on a microcontroller, a component found in many inexpensive electronic devices.
“The proposed insect-inspired navigation strategy is an important step on the way to applying tiny autonomous robots in the real world,” said Guido de Croon, Full Professor in bio-inspired drones and co-author of the article. While this strategy does not generate a map and only enables the robot to return to its starting point, it suffices for many applications. For example, in warehouses or greenhouses, drones could gather data and return to a base station for post-processing, eliminating the need for complex navigation systems.
The practical applications of this technology are vast, from navigation complex labs, in warehouses to detecting gas leaks in industrial sites. This insect-inspired navigation strategy marks a significant advancement towards making tiny autonomous robots a viable and transformative tool in various fields.
For more information, read the full paper “Visual Route-following for Tiny Autonomous Robots” in Science Robotics.
