Swarm Robot Collaborative Mapping — PatSnap Eureka
Swarm Robot Collaborative Mapping 2026
Decentralized SLAM, visual-inertial-UWB fusion, and semantic mapping are converging into deployable multi-robot frameworks. This dataset spans 60+ patent and literature records from 2012 to 2026.
From Single-Robot SLAM to Scalable Swarm Architectures
Swarm robot collaborative mapping combines decentralized SLAM, multi-agent coordination, and distributed data fusion to let groups of robots jointly build spatial models of unknown environments. The field has transitioned from centralized multi-robot systems toward truly swarm-like architectures that provide scalability, flexibility, and fault tolerance—properties absent from single-robot SLAM.
Foundational SLAM techniques are maturing into scalable, fault-tolerant multi-robot frameworks. Visual-inertial odometry fused with UWB ranging has become the de facto localization stack for GPS-denied environments, with multiple independent 2020–2022 results demonstrating centimeter-level relative state estimation in aerial swarms.
Semantic and metric-semantic mapping represents a higher-order capability emerging in the dataset. Kimera-Multi (2021) produces dense 3D mesh models with per-face semantic labels, while work on distributed semantic label reconciliation through multiway matching addresses online consistency across independently-operating robots.
In this dataset, South Korea holds the largest share of named-assignee patent activity, with ETRI and Korea Institute of Industrial Technology accounting for the foundational filings. India shows the highest volume of recent pending applications (4 filings, 2024–2026) in retrieved records, signaling a rising academic-institutional patent pipeline.
Publication Activity and Jurisdictional Distribution
Records in this dataset span 2012–2026, with a clear concentration in the 2019–2023 window indicating an active growth phase. Patent filings cluster in South Korea, the United States, and India, while the literature corpus is strongly international.
Named Assignee Patent Filings by Organization in This Dataset
In this dataset, Korea Institute of Industrial Technology and ETRI together account for the largest share of named-assignee filings, with 2 and 2 records respectively, followed by single filings from Seoul National University, LG Electronics, and Indian academic institutions.
↗ Click bars to explorePublication Records by Period in This Dataset
In this dataset, the 2019–2023 window accounts for the largest concentration of records, reflecting the active growth phase of collaborative swarm SLAM research, while 2024–2026 shows a smaller but emerging set of filings.
↗ Click bars to exploreKey Use Cases for Swarm Collaborative Mapping
Records in this dataset identify multiple deployment domains for swarm collaborative mapping, spanning emergency response, environmental monitoring, defense, space exploration, and consumer robotics. Each domain places distinct requirements on communication, localization, and mapping architecture.
Search and Rescue Missions
A 2020 study demonstrated a UAV-humanoid team building occupancy grid maps of post-disaster sites without GPS. A 2021 study addressed cave mapping under communication and power constraints using multiple aerial robots. A 2019 study demonstrated floor-by-floor mapping by a miniature swarm teaming with wall-climbing units, without external infrastructure.
Disaster ResponseDefense and Contested Environments
ETRI’s 2021 US patent describes battlefield situational mapping via mesh-networked autonomous driving robots collecting visual and spatial data. A 2021 framework integrates SAR and EO/IR mapping with route re-planning for drone swarm missions in hostile environments. These systems require bandwidth-adaptive, fault-resilient map sharing under active interference.
Defense MappingPlanetary and Space Exploration
A 2019 study proposes spacecraft swarms for complete surface mapping during flyby missions using attitude control. The 2021 TEAM system (Trilateration for Exploration and Mapping with Robotic Networks) is explicitly motivated by lunar exploration scenarios. These applications require fully infrastructure-free localization and fault-tolerant coordination.
Space MappingConsumer Robotics Facility Mapping
LG Electronics’ 2025 IN patent filing covers mechanisms for mapping a work environment shared across a plurality of robots, targeting surface cleaning and facility management applications. The earliest patent in this dataset, filed by Korea Institute of Industrial Technology in 2012, covers collaborative map-guided sweeping by robot swarms. These represent the clearest near-term commercial revenue models in the dataset.
Consumer RoboticsKey Patent Assignees in Swarm Collaborative Mapping (Retrieved Records)
Among the 8 named patent assignees in retrieved records, South Korean organizations account for the largest share of filings in this dataset, with ETRI holding 2 records (KR 2013, US 2021) and Korea Institute of Industrial Technology holding 2 US records (2012, 2014). Indian academic institutions collectively filed 3 pending applications in 2024–2026 in this dataset.
Top Assignees by Filing Count in Retrieved Records (Dataset Snapshot)
↗ Click bars to exploreElectronics and Telecommunications Research Institute
ETRI holds 2 filings in this dataset: a 2013 KR patent on SLAM-based area partitioning for collective intelligent robots, and a 2021 US patent on multi-agent manned-unmanned battlefield collaboration using mesh-networked autonomous driving robots. The 2013 KR filing introduced area partitioning and global map fusion using selective matching to reduce merge time. The 2021 US patent is active and extends ETRI’s work into defense-oriented spatial awareness.
South KoreaKorea Institute of Industrial Technology
Korea Institute of Industrial Technology holds 2 US patents in this dataset, filed in 2012 and 2014, establishing the core concept of robots generating local maps, sharing them, and fusing into a global object map for collaborative sweeping tasks. These are the earliest named-assignee filings in this dataset and cover swarm robot sweeping methods with collaborative map-guided execution. Both patents are in the US jurisdiction.
South KoreaNew Frontiers in Swarm Collaborative Mapping (2022–2026)
The most recent records in this dataset (2022–2026) point to six gaining directions: machine learning-driven exploration, heterogeneous aerial-ground teaming, range-based georeferencing, consumer domestic mapping, generative AI for reconstruction, and communication-degraded collaborative SLAM.
Machine Learning–Driven Exploration Policies
Multi-Robot Active Mapping via Neural Bipartite Graph Matching (2022) uses a multiplex graph neural network to compute neural distances between robots and frontier nodes, solving goal assignment as bipartite matching to maximize long-term map coverage. Swarm Cooperative Navigation Using Centralized Training and Decentralized Execution (2023) applies multi-agent reinforcement learning to replace hand-coded frontier assignment with learned long-horizon planning policies. These approaches shift swarm coordination from rule-based to data-driven architectures.
Heterogeneous Aerial-Ground Teaming
VIO-UWB-Based Collaborative Localization and Dense Scene Reconstruction (2022) solves full relative pose estimation for aerial-ground heterogeneous teams using UWB ranging and VIO, with LiDAR onboard ground robots for full 3D reconstruction. Collaborative Localization of Aerial and Ground Mobile Robots through Orthomosaic Map (2020) demonstrates UAV aerial overviews combined with ground robot detail mapping. These systems exploit complementary sensing and mobility modalities across platform types.
Decentralized vs. Centralized Multi-Robot SLAM
Click any row to explore further.
| Dimension | Decentralized SLAM | Centralized SLAM |
|---|---|---|
| Architecture | Each robot maintains local map and pose; peer-to-peer data exchange | Robots send data to a central server handling global optimization |
| Scalability | Scales with swarm size; no single bottleneck | Central server becomes bottleneck at large swarm sizes |
| Fault Tolerance | High; loss of one robot does not collapse the system | Low; central server failure disrupts all mapping |
| Communication Load | Compact maplets or incremental exchanges reduce bandwidth (e.g. Maplets 2020) | All raw or processed data routed through central server; higher load |
| Localization Accuracy | Centimeter-level via visual-inertial-UWB fusion (Omni-Swarm 2022, Decentralized VI-UWB 2020) | C2VIR-SLAM (2022) offloads global optimization to server while preserving agent autonomy onboard |
| Semantic Capability | Kimera-Multi (2021) produces dense 3D mesh with per-face semantic labels in a distributed manner | COVINS (2021) uses a central collaborative server for visual-inertial SLAM |
| GPS-Denied Suitability | Strong; demonstrated in aerial swarm flight experiments without GPS | Requires server connectivity; less suited to fully GPS-denied environments |
| Representative Dataset Records | Swarm SLAM (2021), Maplets (2020), Kimera-Multi (2021), Omni-Swarm (2022) | C2VIR-SLAM (2022), COVINS (2021) |
Frequently Asked Questions: Swarm Robot Collaborative Mapping
Swarm robot collaborative mapping encompasses the methods, architectures, and algorithms by which decentralized groups of robots jointly construct spatial representations of unknown environments, combining simultaneous localization and mapping (SLAM), multi-agent coordination, and distributed data fusion. It is distinguished from single-robot SLAM by its scalability, flexibility, and fault tolerance properties.
Visual-inertial-UWB fusion is the dominant approach in this dataset. Omni-Swarm (2022) combines stereo wide field-of-view cameras, UWB sensors, and visual-inertial odometry to achieve centimeter-level relative state estimation. Decentralized Visual-Inertial-UWB Fusion (2020) validated centimeter-level precision in flight experiments, outperforming standalone UWB or vision methods.
In this dataset, Korea Institute of Industrial Technology and Electronics and Telecommunications Research Institute (ETRI) each hold 2 named patent records. Korean organizations account for the foundational filings in KR and US jurisdictions. LG Electronics, Seoul National University R&DB Foundation, and three Indian academic institutions each hold 1 filing in this dataset.
Records in this dataset identify search and rescue, defense and contested environments, environmental monitoring and agriculture, infrastructure inspection, planetary and space exploration, consumer robotics and facility management, and astrophysics (robotic fiber positioners in telescope arrays) as application domains.
Semantic mapping annotates spatial maps with object categories and place labels. Kimera-Multi (2021) produces dense 3D mesh models with per-face semantic labels in a distributed manner. Multi-Robot Distributed Semantic Mapping (2021) addresses online multi-robot semantic label reconciliation using multiway matching, without requiring shared pre-trained classifiers across robots.
Six directions are gaining traction in this dataset’s most recent records: machine learning-driven exploration and mapping policies using MARL and graph neural networks; heterogeneous aerial-ground teaming; range-based infrastructure-free georeferencing; consumer and domestic collaborative mapping (e.g., LG Electronics 2025); generative AI and deep learning for map reconstruction (GLA University 2025); and collaborative SLAM in communication-degraded or contested environments.
Data and insights on this page are based on a limited patent and literature dataset and are for reference only. Figures may not represent the complete technology landscape.