Ultra-Lightweight Collaborative Mapping for Robot Swarms

TL;DR

This work tackles infrastructure-free SLAM for robot swarms by delivering a fully distributed onboard C-SLAM that operates on ultra-constrained hardware. It combines sparse depth sensing, scan-matching loop closures, and a cascaded pose-graph optimization executed locally on each agent, with inter-drone constraints derived through a robust, anchor-based scheme and a token-based UWB communication protocol to scale to hundreds of robots. The approach significantly reduces data traffic and hardware costs (approximately $20 per unit) while achieving mapping accuracy below 30 cm on centimeter-scale drones, outperforming many high-end systems in resource efficiency. Practical validation includes controlled maze experiments, real-world indoor mapping, scalability up to 200 agents, and onboard runtime/energy analyses, demonstrating robust, scalable, and open-source autonomous collaborative mapping for low-power swarms.

Abstract

A key requirement in robotics is the ability to simultaneously self-localize and map a previously unknown environment, relying primarily on onboard sensing and computation. Achieving fully onboard accurate simultaneous localization and mapping (SLAM) is feasible for high-end robotic platforms, whereas small and inexpensive robots face challenges due to constrained hardware, therefore frequently resorting to external infrastructure for sensing and computation. The challenge is further exacerbated in swarms of robots, where coordination, scalability, and latency are crucial concerns. This work introduces a decentralized and lightweight collaborative SLAM approach that enables mapping on virtually any robot, even those equipped with low-cost hardware and only 1.5 MB of memory, including miniaturized insect-size devices. Moreover, the proposed solution supports large swarm formations with the capability to coordinate hundreds of agents. To substantiate our claims, we have successfully implemented collaborative SLAM on centimeter-size drones weighing 46 g. Remarkably, we achieve a mapping accuracy below 30 cm, a result comparable to high-end state-of-the-art solutions while reducing the cost, memory, and computation requirements by two orders of magnitude. Our approach is innovative in three main aspects. First, it enables onboard infrastructure-less collaborative mapping with a lightweight and cost-effective (\$20) solution in terms of sensing and computation. Second, we optimize the data traffic within the swarm to support hundreds of cooperative agents using standard wireless protocols such as ultra-wideband (UWB), Bluetooth, or WiFi. Last, we implement a distributed swarm coordination policy to decrease mapping latency and enhance accuracy.

Figures (13)

• Figure 1: Illustration of a swarm of miniaturized UAVs using our collaborative SLAM system to map a real environment.
• Figure 2: The C-SLAM scheme. (A) Illustration of how a drone acquires a scan once reaching a texture-rich location. (B) Composite visualization of individual pose graphs and how they are connected through inter-drone loop closure edges. (C) The cascaded distributed SLAM optimization, illustrating the graph optimized by each drone.
• Figure 3: An example graph consisted of 6 internal poses and one external pose, showing how an external pose is transformed into an internal constraint.
• Figure 4: The communication protocol, supporting a variable number of swarm agents while minimizing data traffic.
• Figure 5: The state machine of the exploration algorithm illustrating the behavior in each state and the transition conditions.