Collaborative Simultaneous Localization and Mapping in Multi-Robot Systems

Collaborative SLAM extends the traditional single-robot mapping paradigm to teams of mobile agents that simultaneously explore and build a shared representation of unknown environments. By sharing sensor data, pose estimates and partial maps, multi-robot systems achieve greater spatial coverage, improved accuracy through redundancy and enhanced robustness against individual failures. Architectures range from centralised servers that aggregate and optimise all incoming measurements to fully decentralised networks in which each agent performs local computation and exchanges concise summaries with its neighbours. Key challenges include synchronising heterogeneous sensor modalities, associating features across independent trajectories, mitigating communication latency in bandwidth-constrained settings and merging partial maps while preserving consistent scale and coordinate alignment. Techniques such as loop closure detection, distributed data fusion and inter-agent ranging have emerged to correct drift and facilitate global map convergence. Applications span search and rescue in disaster zones, agricultural monitoring with aerial and ground vehicles, inspection of industrial sites and autonomous warehouse logistics. Advances in lightweight onboard processing, wireless communications and robust optimisation have propelled collaborative SLAM towards real-time deployment in complex, large-scale environments.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent surveys of distributed localisation algorithms have provided comprehensive taxonomies of approaches enabling robots to estimate their relative positions without relying on centralised computation. These works classify methods by measurement type—distance-based, bearing-based and multi-measurement fusion—and highlight trade-offs among communication overhead, computational complexity and system robustness. In parallel, reviews of data fusion strategies for collaborative SLAM have detailed architectural frameworks—centralised, decentralised and hybrid—and evaluated machine-learning methods for feature matching and map merging. More recent experimental studies have introduced ultra-wideband ranging networks to support scalable relative pose estimation in three-dimensional environments. By equipping agents with multiple UWB antennas and employing learned bias correction, these systems achieve sub-metre localisation accuracy while reducing communication demands. Together, these contributions underpin new multi-robot deployments that combine inertial and range measurements, robust loop detection and distributed optimisation to deliver cohesive and accurate global maps in challenging real-world scenarios.

Collaborative Simultaneous Localization and Mapping in Multi-Robot Systems publication trend

The graph below shows the total number of articles in collaborative simultaneous localization and mapping in multi-robot systems across all publications each year (not limited to Nature Index journals).

Technical terms

Simultaneous Localization and Mapping (SLAM): The process by which a mobile agent builds a map of an unknown environment while simultaneously determining its own pose within that map.

Distributed SLAM: A multi-robot SLAM architecture in which robots perform local mapping and share information with peers in a decentralised manner.

Loop Closure: A technique for recognising previously visited locations, enabling correction of accumulated drift in the map and pose estimates.

Data Fusion: The integration of heterogeneous sensor measurements or partial maps into a cohesive global representation.

Ultra-Wideband (UWB) Ranging: A radio-based measurement method that provides high-resolution inter-agent distance estimates for relative localisation.

Visual-Inertial Odometry (VIO): A method combining camera images and inertial sensor data to estimate the motion trajectory of a robot over time.

References

1. Distributed Relative Localization Algorithms for Multi-Robot Networks: A Survey. Sensors (2023).
2. Overview of Multi-Robot Collaborative SLAM from the Perspective of Data Fusion. Machines (2023).
3. MURP: Multi-Agent Ultra-Wideband Relative Pose Estimation With Constrained Communications in 3D Environments. IEEE Robotics and Automation Letters (2024).