Autonomous Shuttle Operated by Grab Reportedly Nearly Struck a Schoolgirl at a Crossing in Punggol, Singapore

A Grab autonomous vehicle undergoing testing in Punggol, Singapore, experienced a near-miss with a schoolgirl at a pedestrian crossing, prompting the Land Transport Authority to suspend testing at similar junctions.

A Grab-operated autonomous shuttle reportedly came close to a schoolgirl crossing a road in Punggol, Singapore. The transport regulator said the vehicle had tracked the pedestrians and begun braking and steering away before a safety operator took control. Grab was directed to pause testing at such junctions pending investigation. The vehicle was testing a route for an on-demand service.

Source: AI Incident Database

Risk classification

  • Primary risk domain: 7 AI system safety, failures, & limitations
  • Primary risk subdomain: 7.3 Lack of capability or robustness

The autonomous vehicle failed to navigate a discretionary left turn safely around pedestrians, requiring human safety operator intervention to prevent a collision.

Causal factors

  • Entity: AI
  • Intent: Unintentional
  • Timing: Post-deployment

The risk was caused by the autonomous vehicle's navigation decisions during a public road test, which was an unexpected outcome of its operational goals.

EU AI Act risk tier

  • Risk tier: 2 High Risk

High Risk: The system is an autonomous vehicle operating in public transport and critical road infrastructure, which has significant implications for public safety.

AI system and alleged parties

  • AI system: WeRide autonomous vehicle (WeRide)
  • AI purpose: Autonomous Driving
  • Behaviour type: Autonomous
  • Alleged developer: WeRide, Automated driving system developers
  • Alleged deployer: Grab, Autonomous vehicle fleet operators, Automated driving system deployers
  • Alleged harmed parties: Women and girls, Road users, Pedestrians, Minors, Girls

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

  • Physical: direct Negligible, indirect Negligible
  • Infrastructure: direct Negligible, indirect Negligible
  • Property: direct Negligible, indirect Negligible
  • Financial: direct Negligible, indirect Negligible
  • Environmental: direct Negligible, indirect Negligible
  • Malicious content: direct Negligible, indirect Negligible
  • Differential treatment: direct Negligible, indirect Negligible
  • Civil rights: direct Negligible, indirect Negligible
  • Democracy: direct Negligible, indirect Negligible
  • Privacy: direct Negligible, indirect Negligible
  • Psychological: direct Negligible, indirect Negligible
  • Epistemic: direct Negligible, indirect Negligible
  • Child sexual exploitation and abuse: direct Negligible, indirect Negligible

People affected

  • Occurrences reported: 1
  • People reportedly exposed: 2

Potential causes

Management

  • Aggressive Expansion Timeline: Focus on expanding AV services in Punggol potentially rushed safety reviews.
  • Inadequate Joint Risk Review: Grab and WeRide did not fully address discretionary turn hazards.

Technology

  • Delayed Autonomous Braking: AV braking was insufficient, requiring safety operator intervention.
  • Imperfect Motion Prediction: AV path planning failed to anticipate the child sudden movement.

Data Inputs

  • Misinterpreted Pedestrian State: System classified pedestrians as stationary instead of intending to cross.

Human Factors

  • Delayed Operator Intervention: Onboard operator intervened late during the discretionary left turn.
  • Sudden Pedestrian Action: Child stepped onto the road suddenly after waiting for a few seconds.

Process and Methods

  • Unadjusted Junction Response: Testing protocols did not optimize AV response for discretionary turns.

Regulatory Environment

  • Permissive Testing Approval: LTA permitted testing at busy junctions before safety rules were finalized.

Information quality

  • Classification confidence: High
  • Reason for confidence: The reports provide consistent, detailed accounts of the incident from official sources (LTA) and video evidence, clearly outlining the actions of the AV, the pedestrians, and the safety operator.

An autonomous vehicle undergoing public testing in Singapore experienced a near-miss with a pedestrian, resolved by safety driver intervention. The incident represents a localized public safety concern with negligible national security implications.

  • Overall national security impact: Negligible
  • Scope: Single nation
  • Primary target: Singapore
  • Alleged perpetrator: Unknown

Threat characteristics

  • Imminence: Long-term. The incident represents a localized testing safety issue rather than an active national security crisis.
  • Autonomy: Human-supervised. The vehicle operated autonomously but with a human safety driver who successfully intervened to prevent a collision.
  • Novelty: Established threat. Autonomous vehicle near-misses and safety driver interventions are established issues with numerous historical precedents.

Impact by dimension

  • Physical security: Negligible. A localized autonomous vehicle near-miss during public testing. No injuries or infrastructure damage occurred, and the incident was mitigated by a safety driver.
  • Information security: Negligible. No evidence of information warfare, intelligence compromise, or systematic manipulation associated with this localized testing incident.
  • Sovereignty: Negligible. Normal regulatory oversight by the Land Transport Authority was exercised to pause testing, representing standard domestic governance with no threat to sovereignty.
  • Economic security: Negligible. No significant threat to national economic stability or strategic technological competitive advantage, despite temporary suspension of local testing.
  • Societal stability: Negligible. While the incident received social media attention, it did not cause mass surveillance, systemic discrimination, or civil unrest.
Explore in the interactive Incident Tracker