This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A driverless-capable metro train on the Magenta Line in Delhi crashed through a depot wall during a manual maintenance procedure. The incident was attributed by the Delhi Metro Rail Corporation to human error, specifically the failure of maintenance staff to re-engage the braking system after it had been decommissioned for workshop checks. No injuries were reported, and the incident was confirmed to be unrelated to the train's automated signalling or driverless capabilities.
A driverless metro train in Delhi, India crashed during a test run due to faulty brakes.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
The incident involves a physical crash of a driverless-capable transit vehicle, which is analyzed under system safety and operational robustness, even though the automated systems were inactive at the time.
Human
Due to a decision or action made by humans
Unintentional
Due to an unexpected outcome from pursuing a goal
Pre-deployment
Occurring before the AI is deployed
The incident was caused by human error and negligence when maintenance staff failed to re-engage the train's brakes, occurring during the trial and commissioning phase before public deployment.
A driverless-capable metro train in Delhi crashed into a depot wall during manual maintenance due to human error. The autonomous systems were inactive, and the incident represents a standard industrial accident with negligible national security implications.
The AI system's automated driverless and signalling capabilities were not active during the incident, as the train was being operated manually within the maintenance depot. Therefore, the AI system did not exhibit any autonomous behavior during the event.
High Risk: The AI system is designed for autonomous train operations on a public metro line, which constitutes critical transport infrastructure.
The train is a new generation driverless metro train designed to operate autonomously without a driver using advanced Communication Based Train Control (CBTC) signalling technology.
The reports are very clear and consistent that the incident was caused by manual human error during maintenance and was not an AI or signalling failure. Multiple official statements from the DMRC confirm the details of the manual operation.
None. All official investigations concluded it was a manual maintenance error.
Technology
Isolated Braking System
Brakes were decommissioned for manual push-pull shunting.
Lack of Depot Signaling
Maintenance area was not controlled by automated signaling systems.
Gradient Washing Ramp
The washing plant was built on a ramp, enabling rollback without brakes.
Data Inputs
Unverified Checklist Inputs
The checklist was ticked off without physical verification of brake status.
Lack of Brake Status Feedback
No automated alert notified the operator that brakes were still isolated.
Human Factors
Operator Overconfidence
Staff assumed brakes were reset and ticked the checklist without checking.
Failure to Test Brakes
Maintenance staff failed to test brakes before the train left the shed.
Negligent Handover Verification
The receiving operator failed to cross-check the brake status.
Process and Methods
Flawed Handover Protocol
Procedures allowed train handover based on unverified paper checklists.
Inadequate Brake Normalization
Standard procedure failed to enforce physical normalization of isolated brakes.
Management
Inadequate Maintenance Oversight
Depot management failed to audit checklist compliance among staff.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
