Geotab’s Data Connector Predictive Safety Analytics and Safety Benchmarking
Support Document
5 mins to read
Geotab’s Data Connector Predictive Safety Analytics and Safety Benchmarking is an innovative approach that leverages driver behavior data and machine learning to anticipate and prevent safety incidents. Safety Benchmarking offers an objective comparison of fleet safety performance against industry standards.
User Guide
February 2025
Introduction
Geotab gathers fault codes from various sources and stores them in the FaultData entity within MyGeotab. There are two primary sources for these codes:
- Diagnostic Trouble Codes (DTCs): These codes signal potential vehicle problems. While some are legally mandated, others are manufacturer-specific. Common DTC standards include OBDII for light and medium-duty vehicles and J1939 for heavy-duty vehicles.
- GoDevice-Specific Faults: These codes flag issues with the GoDevice itself, or they may combine vehicle data to generate a fault code based on device logic.
Fault monitoring allows customers to understand whether or not a fault is persistent, the lifecycle of a fault, and the history of faults on a vehicle. Future functionality will support determining the chance of the vehicle being towed due to the fault code. This table provides a clearer, more organized view of fault data, making it easier for customers to:
- Identify Persistent Faults: By combining instances of faults and applying logic to determine that a fault cycle has ended, we are able to understand the persistence of a fault and what faults are currently present in the vehicle.
- Troubleshoot efficiently: The table provides a concise summary of fault information, allowing users to quickly diagnose issues and take corrective action.
- Fault Severity (Not yet available): Understand which fault codes most commonly lead to towing events based on the DTC cycles and downtime data.
- Historical Understanding of Vehicle Issues : By understanding the fault lifecycles we are not only able to identify persistent faults, but also a clear historical understanding of the vehicle's issues.
- Analyze fleet-wide trends: Improved data organization enables users to identify patterns and trends in fault data across their entire fleet, leading to better maintenance planning and reduced downtime.
Fault Cycles
Aggregate fault state cycles
Fault cycles are defined by combining the consecutive ignition cycles and faults with the same state. The green boxes represent the ignition cycles without the fault detected. All other colors represent the series of consecutive ignition cycles having faults with the corresponding fault state.
Aggregate Fault Cycles
We combine consecutive fault state cycles for every vehicle to form a fault cycle. The end of the cycle is defined by ignition cycles where faults are not recorded (green box). If we detect enough ignition cycles where the fault is not recorded, we define the cycle as “Not Persistent”.
Any new fault of the same type on the same vehicle will open a new cycle with a new ID for the cycle.
Visualizing Fault Monitoring
Risk of Breakdown — Fault Severity (Beta)
The risk of breakdown model aims to correlate the fault lifecycle with the determined towing events. The following is a visual representation of the model:
Therefore the risk of breakdown model, at this time, is defined as the percentage chance of a vehicle ending in a towing event.
Fault severity is grouped based on the relative chance of tow meaning:
- Low: 0-4%
- Medium: 4-8.5%
- High: 8.5-15%
- Critical: >15%
✱ NOTE: This data is currently in under development and is considered a Beta feature. Users should be aware that this may result in unintended errors and that the feature may change at any time.
Data Schemas
Fault monitoring provides the data required to visualize faults and how long the vehicle has been driving with this issue in both mileage and time. It is recommended to use AnyStatesDateTimeFirstSeen and AnyStatesDateTimeLastSeen as a starting point, as it represents the entire fault cycle. By using the isPersistentCycle column, you can determine which fault codes Geotab continues to view as on.
Columns with Pending, Active, and Confirmed faults provide more detailed information about the lifecycle of the DTC.
FaultMonitoring
Column Name | Column Description |
DeviceId | Unique identifier of an asset in the MyGeotab database. |
SourceId | Unique identifier of the fault source in the MyGeotab database. |
FaultCode | Decoded Fault Code. |
FaultCodeDescription | Description of the Fault Code |
DiagnosticType | Type of the FaultCode (e.g. OBDII, SPN, PID/SID etc.) |
FailureMode | FailureMode of the FaultCode |
FailureModeDescription | FailureMode Description |
Controller | Controller of the FaultCode |
ControllerDescription | Controller Description |
Component | Component of the FaultCode |
AnyStatesDateTimeFirstSeen | First DTCDateTime of the DTC Cycle |
AnyStatesDateTimeLastSeen | Last DTCDateTime of the DTC Cycle |
IsPersistentCycle | Indication whether or not the DTC cycle is considered as persistent/on-going. If the cycle is persistent the last seen DTC is not yet set as the DTC cycle end and can still update. |
PendingDateTimeFirstSeen | DateTime of the first Pending DTC. |
PendingOdometerFirstSeen | Odometer value of the first Pending DTC. |
PendingDateTimeLastSeen | DateTime of the last Pending DTC. |
PendingOdometerLastSeen | Odometer value of the last Pending DTC. |
PendingDuration | Duration where the DTC cycle is in the pending state. |
PendingDistance | Distance driven with DTC cycle in pending state. |
PendingCount | Count of all Pending DTCs inside the cycle. |
ActiveDateTimeFirstSeen | DateTime of the first Active DTC. |
ActiveOdometerFirstSeen | Odometer value of the first Active DTC. |
ActiveDateTimeLastSeen | DateTime of the last Active DTC. |
ActiveOdometerLastSeen | Odometer value of the last Active DTC. |
ActiveDuration | Duration where the DTC cycle is in the active state. |
ActiveDistance | Distance driven with DTC cycle in active state. |
ActiveCount | Count of all Active DTCs inside the cycle. |
ConfirmedDateTimeFirstSeen | DateTime of the first Confirmed DTC. |
ConfirmedOdometerFirstSeen | Odometer value of the first Confirmed DTC. |
ConfirmedDateTimeLastSeen | DateTime of the last Confirmed DTC. |
ConfirmedOdometerLastSeen | Odometer value of the last Confirmed DTC. |
ConfirmedDuration | Duration where the DTC cycle is in the confirmed state. |
ConfirmedDistance | Distance driven with DTC cycle in confirmed state. |
ConfirmedCount | Count of all Confirmed DTCs inside the cycle. |
FaultMonitoring_Daily
Provides the daily log of every fault cycle, between the times of AnyStatesDateTimeFirstSeen and AnyStatesDateTimeLastSeen.
Column Name | Column Description |
Local_Date | Local date of the record in the timezone of the device/asset |
DeviceId | Unique identifier of an asset in the MyGeotab database. |
SourceId | Unique identifier of the fault source in the MyGeotab database. |
FaultCode | Decoded Fault Code. |
FaultCodeDescription | Description of the Fault Code |
DiagnosticType | Type of the FaultCode (e.g. OBDII, SPN, PID/SID etc.) |
FailureMode | FailureMode of the FaultCode |
FailureModeDescription | FailureMode Description |
Controller | Controller of the FaultCode |
ControllerDescription | Controller Description |
Component | Component of the FaultCode |