Autonomous robot coordination

Yes. There are several fleet-management / orchestration platforms for warehouse AMRs that explicitly use “swarm” or decentralized multi-robot coordination, although the term swarm intelligence is used pretty loosely in the industry.

Strongest matches

Tool / vendor	Swarm / decentralized?	Collision & traffic coordination	Fit
Locus Robotics — LocusONE	Yes — “swarm AI”	Dynamic routing and coordination of multiple robot types	⭐⭐⭐⭐⭐
FARobot — SWARM CORE	Yes	Multi-robot traffic control, route planning, decentralized/P2P communication, rerouting	⭐⭐⭐⭐⭐
MOOE Robotics — RCS	Yes — explicitly decentralized	Collision avoidance, dynamic traffic optimization, path replanning	⭐⭐⭐⭐
KUKA.AMR Fleet	AI-based, but not marketed primarily as swarm	Real-time traffic-aware replanning and obstacle avoidance	⭐⭐⭐⭐
Rayantra	Yes — swarm orchestration	Conflict-free swarm planning and shared-world-model coordination	⭐⭐⭐⭐

LocusONE is particularly interesting if you're looking for an actual commercial warehouse deployment rather than a research project. Locus describes its platform as using proprietary swarm AI to coordinate its robot fleet, dynamically route robots, and manage tasks in real time.locusrobotics.com

FARobot's SWARM CORE is an even closer conceptual match to what you're describing. It coordinates robots in route, traffic, and task planning; supports heterogeneous robot fleets; and specifically describes multi-robot traffic control designed for collision-free operation. It also describes decentralized/P2P communication for dynamic decision-making.www.farobottech.com

MOOE's RCS makes the decentralized claim particularly explicitly: it says individual robots make local decisions while coordinating with the fleet, with collision avoidance, dynamic traffic optimization, path planning/replanning, and deadlock-related safety mechanisms.www.farobottech.com+1

One important distinction

There are actually two different problems that get lumped together as "preventing robots from crashing":

1. Local collision avoidance — robot A sees robot B and brakes/steers around it.
2. Fleet-level traffic coordination — the system predicts that A and B will meet at an intersection 10 seconds from now, assigns right-of-way or changes their routes, and prevents the conflict altogether.

The second is where swarm/multi-agent intelligence becomes especially useful. For example, recent MIT/Symbotic work uses reinforcement learning to determine which robots should receive priority as congestion develops, rather than merely reacting after robots encounter one another.news.mit.edu

There are also research/prototype systems using particle-swarm optimization, artificial potential fields, ORCA, conflict-based search, auctions, and decentralized negotiation for exactly this problem.link.springer.com

My take

If your requirement is:

“We already have warehouse robots and want software that coordinates the fleet so they don't deadlock or collide.”

I'd investigate FARobot SWARM CORE and LocusONE first, then compare them with conventional centralized fleet managers such as KUKA.AMR.

The critical questions to ask vendors are not simply “Do you use swarm intelligence?” Ask:

• Does each robot make decisions locally, or is there a central traffic planner?
• How are head-on conflicts resolved?
• How are intersection reservations/right-of-way handled?
• Can the system prevent deadlocks, not just collisions?
• What happens when Wi-Fi/network connectivity is lost?
• Does it coordinate robots from different manufacturers?
• Does it replan in real time when a robot stops unexpectedly?
• Is collision avoidance handled at the fleet level and independently on the robot's safety controller?
• What's the maximum demonstrated fleet size and robot density?

That last point is important: fleet orchestration should not be your primary safety layer. Physical LiDAR/vision safety systems and emergency stops should remain independent safeguards.

If you tell me what robots you have (e.g., MiR, Locus, OTTO, Geek+, custom AMRs, forklifts) and roughly how many, I can narrow this down to the 3–5 systems that could actually integrate with your fleet.

Yes. The important distinction is that AMR routing is often split between a WES/WCS and the AMR fleet manager: the WES decides what work should happen and in what priority, while the fleet-control layer handles which robot takes it, traffic, path planning, and rerouting.

Strong candidates for AMR routing

Platform	AMR routing / orchestration	Best fit
Swisslog SynQ + IntraMove	Excellent. AI fleet management dynamically replans routes, handles obstacles, allocates transport tasks, balances loads, and supports VDA 5050 for heterogeneous fleets. SynQ adds order prioritization and higher-level orchestration.www.swisslog.com	Large automated DCs; mixed AMR/automation environments
OMRON FLOW Core	Excellent for fleet-level routing. Provides dynamic traffic control, job assignment, route planning, rerouting, live robot status and charging management. OMRON says it can coordinate fleets of up to 100 OMRON AMRs.robotics.omron.com	AMR-centric warehouses and manufacturing
Honeywell Momentum WES	Excellent for system-level optimization. Continuously optimizes order routing, workload, capacity and congestion while integrating AMRs, conveyors, AS/RS and other automation.automation.honeywell.com	Large, highly automated distribution centers
Dematic WCS/WES + AMRs	Very strong. Dematic describes its AMR software as providing real-time route optimization and workload balancing, with AMRs integrated into broader warehouse automation.www.dematic.com	High-volume fulfillment with conveyors/sortation
Blue Yonder WES / Robotics Hub	Very strong at orchestration. Its WES dynamically assigns work to humans and robots, while Robotics Hub provides a vendor-agnostic integration layer for AMRs and other robotics. Its optimization objectives can include minimizing travel time.blueyonder.com	Enterprise warehouses with mixed labor + robotics
Körber WCS/WES	Strong for heterogeneous automation; particularly attractive when AMRs need to coexist with AS/RS, conveyors and other equipment.	Multi-vendor automation
Locus Robotics LocusOne	Strong specialized AMR fleet orchestration, particularly for goods-to-person/e-commerce fulfillment.	AMR-heavy fulfillment operations

My shortlist

If your specific question is “Which systems actually optimize AMR routes and traffic?”, rather than simply integrating with AMRs, I'd put these at the top:

1. Swisslog SynQ + IntraMove — strongest combination of WES/WCS orchestration and dynamic AMR routing.
2. OMRON FLOW Core — particularly strong if fleet traffic control and dynamic rerouting are the core requirement.
3. Honeywell Momentum WES — strongest candidate when AMRs are one component of a much larger automated DC.
4. Dematic WCS/WES — compelling for Dematic-heavy, high-throughput automation.
5. Blue Yonder WES + Robotics Hub — particularly interesting for mixed human/robot operations and multi-vendor robotics.

One key procurement point: don't evaluate only the WCS name. Ask vendors to demonstrate a live scenario with 50–200 AMRs where an aisle becomes blocked, a high-priority order arrives, a robot's battery drops, and several robots converge on the same intersection. The meaningful differentiators are dynamic rerouting, deadlock prevention, traffic/intersection management, task allocation, fleet balancing, and optimization across robots + other automation.

If you tell me whether you're evaluating a specific AMR fleet (Geek+, Locus, MiR, OMRON, etc.) or a warehouse software replacement, I can narrow this to a vendor-by-vendor comparison of routing algorithms, multi-vendor support, VDA 5050, WMS integration, and estimated enterprise fit.