Skip to main contentOverview
Challenge: Managing a fleet of autonomous systems requires tracking health status, detecting anomalies early, and identifying maintenance needs before failures occur across all units.
Solution: Heex enables fleet-wide health tracking and anomaly detection through configured triggers. You define the conditions that matter for your operation, and Heex notifies you when those conditions occur across any robot in your fleet.
Key Benefits:
- Centralized trigger configuration for entire fleet
- Smart notifications based on your defined conditions
- Consistent monitoring patterns across all systems
- Fleet-wide event tracking and analysis
- Proactive maintenance through condition-based alerts
Use Case: Warehouse Robot Fleet
Scenario
A logistics company operates 50 autonomous mobile robots (AMRs) in a 200,000 sq ft warehouse. The operations team needs to:
- Get notified when battery conditions require attention across any robot
- Be alerted when robots get stuck or experience navigation issues
- Receive alerts when performance degrades on specific units
- Be notified of maintenance needs before failures occur
Traditional Approach (Without Heex)
- Each robot logs to local storage
- No centralized notification system
- Operators discover issues reactively when robots stop working
- Manual inspection required to diagnose problems
- No systematic approach to condition monitoring
Problems:
- Average downtime: 4 hours per incident
- Unplanned maintenance: 80% of all maintenance
- Data analysis: Manual, time-consuming, incomplete
Heex Approach
- All robots send data to central Heex workspace
- Configure triggers once, apply to entire fleet
- Receive notifications when conditions match across any robot
- Track events and patterns across the fleet
- Proactive maintenance through condition-based triggers
Results:
- Average downtime: 30 minutes per incident (-87%)
- Unplanned maintenance: 30% of all maintenance (-62%)
- Event tracking: Automated, centralized, actionable
Implementation Approach
Step 1: Define Your Monitoring Strategy
Identify the key conditions you want to monitor across all robots in your fleet. For a warehouse AMR fleet, this typically includes:
- Battery Health: Voltage, temperature, charge cycles, aging patterns
- Navigation Health: Position tracking, stuck detection, path efficiency
- Performance Tracking: Task completion times, idle periods, error rates
Each robot needs a datasource containing the signals you want to monitor. For a 50-robot fleet:
- Create 50 datasources (one per robot)
- Tag each with identifying labels (robot ID, zone, model)
- Use standardized signal names across all datasources
Consistent signal naming ensures triggers can be applied uniformly to the entire fleet.
Step 3: Create Fleet-Wide Triggers
Design triggers that define the specific conditions warranting notification. Examples for warehouse AMRs:
Critical Battery Alert: When battery voltage drops below 11.5V, generate a high-priority notification. This gives operators time to send the robot to charge before it stops mid-task.
Battery Overheating: When battery temperature exceeds 60°C, immediately notify operations with critical priority to prevent damage or fire risk.
Predictive Maintenance - Battery Aging: When a robot has accumulated over 800 charge cycles AND voltage is below 12.0V, trigger a maintenance notification. This combination indicates the battery should be replaced before it fails unexpectedly.
Navigation Issue - Robot Stuck: When a robot remains in the same position for more than 5 minutes while not in charging or idle mode, notify operators of potential navigation problems.
Performance Degradation: When task completion time consistently exceeds the fleet average by 50%, generate an investigation alert to identify mechanical or software issues.
Step 4: Apply Triggers to Systems
Assign configured triggers to the relevant systems in your fleet:
- Apply universally for common conditions (battery, critical alerts)
- Target specific zones or robot types using labels
- Roll out gradually for validation (test on 5-10% first)
This centralized configuration means you define monitoring logic once and apply it consistently across your entire operation.
Fleet Insights from Events
When triggers match their configured conditions, Heex generates events centralized in your workspace. Over time, your event history reveals patterns across your fleet:
Battery Health Trends: Review which robots generate the most battery-related events. If certain units consistently hit low battery warnings earlier than others, they may need battery replacement or workload rebalancing.
Zone-Specific Issues: Events tagged by warehouse zone can reveal environmental patterns. For example, if robots in the shipping zone consistently overheat while those in picking zones don’t, it may indicate HVAC or ventilation issues.
Predictive Maintenance Queue: Events labeled for maintenance action create a natural queue of robots requiring service. Filter for open maintenance events and schedule interventions during planned downtime windows.
Performance Outliers: When performance degradation triggers fire for specific robots, compare their event patterns to fleet average. Repeated performance alerts may indicate mechanical wear, software bugs, or configuration issues requiring investigation.
Next Steps
Learn More About Heex Concepts:
Apply Best Practices:
Explore Related Use Cases:
Reference Documentation:
