A warehouse running 28 AMRs hit 74% fleet availability during peak season — not because robots failed suddenly, but because battery degradation across 11 units accumulated undetected over three months, wheel wear on high-traffic routes went uninspected, and two LiDAR sensors drifted without calibration. The fleet that was supposed to process 1,400 picks per hour was delivering 980. The fix was not new hardware — it was a structured maintenance programme in OxMaint that registered every robot as an individual asset, automated PM scheduling by component and runtime hours, and flagged degradation before it reached the floor. Fleet availability returned to 96% within 60 days. Every warehouse running AMRs faces this trajectory without a CMMS purpose-built for fleet-scale operations — book a demo to see how OxMaint manages your fleet.
Warehouse Robotics / Fleet Maintenance
AMR Fleet Predictive Maintenance for Warehouse Operations Using CMMS
From 74% to 96% fleet availability. How CMMS-driven predictive maintenance transforms AMR operations — with automated work orders, robot-level PM scheduling, and fleet health analytics built for warehouse scale.
Fleet Availability — Before vs After CMMS
Without structured maintenance
After 60-day OxMaint programme
Industry target for AMR fleets
Based on structured PM across battery, drive, sensor, and software domains
The Four Failure Modes That Drive 80% of AMR Downtime
Pareto analysis of AMR fleet work order data across warehouse deployments consistently identifies four root causes behind the vast majority of unplanned stoppages. Understanding these failure modes — and their detection windows — is the foundation of an effective predictive maintenance programme.
01
Battery Degradation
Accounts for ~35% of fleet stoppages
Lithium-ion AMR batteries last 2,000–3,000 full charge cycles — roughly 3–5 years in warehouse service. Capacity decline is gradual and invisible without tracking. A battery at 70% capacity returns robots to charging mid-route, reduces shift coverage, and causes navigation errors when voltage drops cause sensor brownouts. Battery state-of-health tracking per unit, not per fleet average, is the only way to catch individual units degrading faster than the fleet baseline.
Detection window: 6–12 weeks before stoppage
02
Drive Wheel and Caster Wear
Accounts for ~25% of fleet stoppages
Drive wheels wear at rates determined by payload weight, floor surface, and route intensity. High-traffic routes — inbound dock to staging, staging to pick zones — create uneven wear profiles across a fleet. Worn wheels cause odometry errors, which manifest as navigation drift, increased path correction frequency, and eventual localisation failures. Wheel diameter reduction of just 3–5mm causes measurable odometry deviation that compounds with every kilometre travelled.
Detection window: 4–8 weeks before navigation failure
03
Navigation Sensor Contamination
Accounts for ~22% of fleet stoppages
LiDAR sensors and cameras accumulate dust, pallet wrap debris, and condensation — particularly in high-throughput zones near receiving docks and packaging lines. Contaminated sensors reduce detection range, increase obstacle avoidance false triggers, and cause emergency stops in clear aisles. Sensor cleaning is a simple 5-minute task that prevents a 45-minute stoppage, but without a scheduled PM work order, it does not happen consistently across a large fleet.
Detection window: 1–3 weeks (visible in false stop rate data)
04
Software and Firmware Faults
Accounts for ~18% of fleet stoppages
Navigation map drift, fleet manager software bugs, and firmware version mismatches between robots cause coordination failures, task assignment errors, and robot-to-robot collision near-misses. Firmware updates are often deferred because the maintenance window is tight and the risk of update failure is real. Without CMMS tracking of firmware versions per unit and structured update scheduling, fleets drift into mixed-version states that create unpredictable interaction failures under load.
Detection window: Visible in error log trending before impact
Register Every Robot. Track Every Failure Mode. Prevent the Stoppage.
OxMaint registers each AMR as an individual asset with its own PM schedule, battery health tracking, work order history, and spare parts allocation — giving you fleet-level visibility and robot-level precision across every maintenance task.
AMR Health Scorecard — What to Track Per Robot in Your CMMS
Each AMR in your fleet should have a health profile that captures the six indicators most predictive of imminent failure. Tracking these at the individual robot level — not the fleet average — is what separates a maintenance programme that prevents failures from one that reacts to them.
Battery State of Health
Above 85% capacity
75–85% capacity
Below 75% — schedule replacement
Auto work order at 80%
Drive Wheel Diameter
Within 2mm of nominal
2–4mm reduction
Over 4mm — replace before next shift
PM inspection every 250hrs
False Stop Rate
Under 2 per shift
2–5 per shift
Over 5 — sensor cleaning/check
Alert from fleet manager API
Motor Temperature
Under 55°C at full load
55–65°C
Over 65°C — inspect immediately
Real-time alert + work order
Navigation Map Accuracy
Under 5% deviation events/day
5–10% deviation events/day
Over 10% — map rebuild required
Weekly PM trigger
Firmware Version Status
Current or 1 version behind
2 versions behind
3+ behind — update in next window
Monthly software PM task
How OxMaint Connects to Your AMR Fleet Manager
The maintenance intelligence already exists in your fleet management platform — Locus Robotics, MiR Fleet, Geek+, Fetch/Zebra, 6 River Systems. The gap is that it stays there, disconnected from the CMMS that manages your work orders, parts, and technicians. OxMaint closes that gap.
1
Fleet Manager Telemetry
Battery SOC, motor temperatures, error codes, cycle counts, and false stop rates stream from your AMR fleet manager platform via REST API or MQTT broker — in real time, per robot.
2
OxMaint Threshold Evaluation
Each data point is evaluated against configurable thresholds set per robot model and operating environment. Thresholds are calibrated over the first 30 days and refined continuously as fleet data accumulates.
3
Automatic Work Order Creation
When a threshold is crossed, OxMaint creates a work order pre-populated with robot ID, fault type, last known location, maintenance history, and required spare parts — assigned to the on-shift technician with the relevant skill certification.
4
Maintenance Execution and Feedback
Technician completes the work order on mobile, logs findings and parts used. Completion data feeds back into the robot's health profile, updating MTBF calculations and refining prediction accuracy for the next cycle.
Fleet-Level Analytics That Drive Maintenance Decisions
Individual robot health is necessary but not sufficient. Fleet-level analytics reveal patterns that are invisible at the unit level — which robot models degrade fastest, which routes cause the most wear, which shift patterns correlate with elevated false stop rates. These insights drive PM interval optimisation, route design decisions, and fleet expansion justifications.
MTBF by Robot
Mean time between failures calculated per individual unit — not fleet average. Identifies the 20% of robots generating 80% of reactive work orders. These units get intensified PM intervals before they cascade into mission-critical failures during peak periods.
Battery Degradation Curves
State-of-health trending per battery pack over its operational life. OxMaint projects when each battery will cross the 80% replacement threshold — enabling bulk replacement planning during a single scheduled window rather than reactive swaps every few weeks.
Route-Wear Correlation
Wheel wear rate correlated against route assignment data identifies which facility zones accelerate drivetrain degradation. Maintenance managers use this to rotate robot assignments across routes, extending component life without changing hardware or facility layout.
Spare Parts Consumption Rate
Throughput vs Fleet Health Index
Picks per hour correlated against fleet-wide health score. This chart is the executive-level view that connects maintenance investment to operational output — proving that the maintenance programme protects revenue, not just robots.
Planned vs Unplanned Ratio
Frequently Asked Questions
How does OxMaint integrate with AMR fleet management platforms like MiR, Locus, or Geek+?
OxMaint connects to AMR fleet managers via REST API, MQTT broker, or webhook — receiving telemetry data including battery state of charge, motor temperatures, error codes, runtime hours, and false stop events. Each data stream is mapped to configurable thresholds that trigger work orders automatically when crossed. The integration is bidirectional — maintenance windows and robot-offline status flow back to the fleet manager so it can route tasks around robots scheduled for service.
Sign in to start your OxMaint AMR integration setup.
What spare parts should we stock for a warehouse AMR fleet of 20–30 robots?
For a fleet of 20–30 AMRs, standard safety stock guidance is: drive wheel sets (2 sets per 10 robots), LiDAR modules (1 per 15 robots), battery packs (1 per 8 robots based on 24-month replacement cycles), charging contact assemblies (1 per 8 robots), bumper sensor assemblies (1 per 10 robots), and caster wheel sets (2 sets per 10 robots). OxMaint tracks actual consumption against these baselines and generates reorder alerts before stock falls below minimum levels — eliminating the parts-unavailability delays that extend MTTR.
Book a demo to configure your AMR spare parts register.
How do you schedule AMR maintenance without taking robots out of service during peak hours?
AMR PM scheduling in OxMaint uses a combination of runtime-hour triggers and shift-pattern awareness — scheduling maintenance tasks to execute during charging periods, shift transitions, or defined off-peak windows. Most AMR preventive maintenance tasks — sensor cleaning, wheel inspection, contact cleaning, software review — can be completed in under 30 minutes per unit during a scheduled charging cycle. OxMaint queues work orders to release at the start of the maintenance window and confirms spare parts availability before the window opens, so technicians execute immediately rather than searching for parts.
At what fleet size does AMR maintenance require a CMMS?
The transition point where spreadsheet-based AMR maintenance fails is typically 12–15 robots. Below that threshold, a single technician can track battery health and inspection schedules manually with reasonable reliability. Above it, battery degradation patterns diverge between units, wheel wear rates vary by route, and firmware version drift creates coordination failures that manual tracking cannot catch in time. Facilities scaling from 10 to 30+ robots consistently report that the maintenance chaos emerges faster than expected — and that implementing a CMMS before reaching that threshold is significantly less disruptive than implementing it during a reliability crisis.
Sign in to start your OxMaint AMR asset registry.
Start Your AMR Maintenance Programme
Your AMR Fleet Is Generating Maintenance Intelligence Right Now. Is Your CMMS Listening?
OxMaint connects to your fleet manager, registers every robot as an individual asset, automates PM scheduling across battery, drive, sensor, and software domains, and tracks fleet health analytics that protect your warehouse throughput — all in one platform.
