← Back to incidents
Knightscope K5 Security Robot Runs Over Toddler at Stanford Shopping Center
HighA 300-pound Knightscope K5 autonomous security robot at Stanford Shopping Center knocked down and ran over a 16-month-old toddler in July 2016, causing minor injuries when its sensors failed to detect the child.
Category
Safety Failure
Industry
Other
Status
Resolved
Date Occurred
Jul 7, 2016
Date Reported
Jul 12, 2016
Jurisdiction
US
AI Provider
Other/Unknown
Model
Knightscope K5
Application Type
agent
Harm Type
physical
People Affected
1
Human Review in Place
No
Litigation Filed
No
autonomous_robotpedestrian_safetychild_injurysensor_failurepublic_spacessecurity_robotcollision_avoidance
Full Description
On July 7, 2016, at Stanford Shopping Center in Palo Alto, California, a Knightscope K5 autonomous security robot struck and ran over a 16-month-old toddler. The incident occurred while the child was walking with his family near the outdoor mall's common areas where the 300-pound, five-foot-tall robot was conducting its routine security patrol. The K5's sensor array, which includes cameras, microphones, and various detection systems designed to identify security threats and navigate around obstacles, failed to detect the small child.
The robot knocked the toddler down and continued moving forward, running over the child before stopping. The incident resulted in scrapes and bruising to the child's leg and foot, but fortunately no broken bones or serious injuries. Emergency medical personnel were called to the scene, and the child was treated for his injuries. The family reported that the robot appeared to have no awareness that it had struck the child and continued its patrol until manually stopped.
Knightscope, founded in 2013 and based in Mountain View, California, markets its K5 robots as autonomous security guards capable of patrolling areas, detecting anomalies, and reporting incidents to human security personnel. The robots are equipped with multiple sensors and are designed to operate independently in public spaces including malls, offices, and parking lots. The company had deployed several K5 units across various locations as part of its growing autonomous security robot program.
Following the incident, Knightscope issued a statement characterizing the event as an 'anomaly' and emphasizing that the robot had operated for thousands of hours without previous incidents. The company stated it was investigating the circumstances and working to understand why the robot's obstacle detection systems failed. However, this incident was not isolated - other K5 robots had experienced notable failures, including one unit that fell into a fountain at a Washington DC office complex in July 2017, and reports of robots knocking over or startling pedestrians at various deployment sites.
The Stanford Shopping Center incident exposed significant gaps in safety protocols and regulatory oversight for autonomous robots operating in public spaces. Unlike autonomous vehicles, security robots face limited safety certification requirements despite operating in close proximity to pedestrians including children. The incident raised questions about whether existing sensor technologies and AI algorithms were sufficiently robust for safe operation around vulnerable populations like small children, elderly individuals, and people with disabilities.
The broader implications of this incident contributed to ongoing discussions about liability, insurance, and safety standards for autonomous systems deployed in public environments. While the child's injuries were relatively minor, the incident demonstrated the potential for serious harm when heavy autonomous systems fail to properly detect and avoid human contact, particularly with children whose movements and physical profiles may not align with standard obstacle detection parameters.
Root Cause
The K5 robot's sensor array and collision avoidance algorithms failed to detect and avoid a small child moving at ground level. The robot's design prioritized detection of adult-sized obstacles and threats rather than comprehensive pedestrian safety including children.
Mitigation Analysis
This incident highlights critical gaps in autonomous robot safety certification and testing protocols. Enhanced sensor redundancy with multiple detection modalities (LIDAR, computer vision, proximity sensors) specifically calibrated for child-height obstacles could have prevented the collision. Mandatory safety testing scenarios including children and comprehensive pedestrian interaction protocols should be required before deployment in public spaces.
Lessons Learned
The incident highlighted critical gaps in safety testing and certification for autonomous robots in public spaces, particularly regarding detection of children and vulnerable populations. It demonstrated the need for comprehensive pedestrian safety protocols and regulatory oversight before deploying autonomous systems in environments with public access.
Sources
Security robot hits toddler at California mall
CNN · Jul 12, 2016 · news
Family says security robot at Stanford Shopping Center hit, injured toddler
ABC7 News · Jul 12, 2016 · news