Incident 68: Security Robot Drowns Itself in a Fountain

Description: A Knightscope K5 security robot ran itself into a water fountain in Washington, DC.

Tools

New ReportNew ReportNew ResponseNew ResponseDiscoverDiscover
Alleged: Knightscope developed and deployed an AI system, which harmed Knightscope.

Incident Stats

Incident ID
68
Report Count
30
Incident Date
2017-07-17
Editors
Sean McGregor

CSET Taxonomy Classifications

Taxonomy Details

Full Description

A robot at an office building in Washington, DC ran itself into a water fountain. The robot, named Knightscope K5, was developed as a security robot that uses facial recognition and a variety of sensors to detect criminals. The reasons the robot fell into the fountain are unclear.

Short Description

A Knightscope K5 security robot ran itself into a water fountain in Washington, DC.

Severity

Negligible

Harm Type

Harm to physical property

AI System Description

The robot developed by the start-up contains autonomous driving and facial recognition abilities

System Developer

Knightscope

Sector of Deployment

Administrative and support service activities

Relevant AI functions

Perception, Cognition, Action

AI Techniques

Facial recognition, environmental sensing

AI Applications

autonomous driving, self-driving vehicle

Location

Washington, D.C.

Named Entities

Knightscope, Knightscope K5, Mountain View, Electronic Privacy Information Center (EPIC), Stanford Shopping Center

Technology Purveyor

Knightscope

Beginning Date

2017-07-17

Ending Date

2017-07-17

Near Miss

Unclear/unknown

Intent

Accident

Lives Lost

No

Data Inputs

Photographs

GMF Taxonomy Classifications

Taxonomy Details

Known AI Goal

Autonomous Drones

Potential AI Technology

Image Segmentation

Potential AI Technical Failure

Generalization Failure

Variants

A "variant" is an incident that shares the same causative factors, produces similar harms, and involves the same intelligent systems as a known AI incident. Rather than index variants as entirely separate incidents, we list variations of incidents under the first similar incident submitted to the database. Unlike other submission types to the incident database, variants are not required to have reporting in evidence external to the Incident Database. Learn more from the research paper.

Similar Incidents

By textual similarity

Did our AI mess up? Flag the unrelated incidents