Incident: Faulty Battery Replacement Causes KiwiBot Fire Incident at UC Berkeley

Published Date: 2018-12-16

Postmortem Analysis
Timeline 1. The software failure incident of the KiwiBot catching fire happened on a Friday afternoon [79391]. 2. Article 79391 was published on 2018-12-16. 3. Estimating the timeline: - The incident occurred on a Friday afternoon, which could be assumed to be the same day as the article's publication. - Therefore, the software failure incident happened on December 14, 2018.
System 1. Faulty battery replacement [79391]
Responsible Organization 1. The software failure incident, where a KiwiBot caught fire due to a faulty battery replacement, was caused by human error during a routine battery replacement [79391].
Impacted Organization 1. Customers and members of the public were impacted by the software failure incident involving the KiwiBot robot catching fire due to a faulty battery replacement [79391].
Software Causes 1. The software cause of the failure incident was human error in which a functioning battery was replaced with a defective one during a routine battery replacement, leading to the battery experiencing thermal runaway [79391].
Non-software Causes 1. Faulty battery replacement leading to thermal runaway [79391]
Impacts 1. The impact of the software failure incident involving the KiwiBot catching fire at UC Berkeley was primarily related to safety concerns and operational disruptions. The faulty battery replacement led to the robot experiencing thermal runaway, resulting in a fire incident in front of the school's student union [79391].
Preventions 1. Implementing thorough quality assurance testing procedures to ensure that all components, including batteries, are functioning correctly before deployment [79391]. 2. Developing and implementing software safeguards to prevent the use of defective components, such as implementing software checks to verify the authenticity and functionality of replacement parts [79391]. 3. Regularly updating and maintaining the software monitoring systems to detect any anomalies or potential issues with the robots' batteries or other components [79391].
Fixes 1. Implementing software to monitor the state of the robots' batteries [79391]
References 1. The Daily Californian [79391] 2. KiwiBot's blog post [79391]

Software Taxonomy of Faults

Category Option Rationale
Recurring unknown (a) The software failure incident having happened again at one_organization: The incident involving the KiwiBot catching fire due to a faulty battery replacement is a unique occurrence attributed to human error within the organization. There is no mention in the article of a similar incident happening again within the same organization or with its products and services [79391]. (b) The software failure incident having happened again at multiple_organization: There is no information in the provided article about a similar incident happening again at other organizations or with their products and services [79391].
Phase (Design/Operation) design (a) The software failure incident in Article 79391 was related to the design phase. The incident was attributed to a faulty battery replacement, which was a result of human error during a routine battery replacement process. The company, KiwiBot, mentioned that a functioning battery was replaced with a defective one, leading to the battery experiencing thermal runaway. As a response, the company addressed the issue by installing software to monitor the state of the robots' batteries, indicating a design-related failure introduced during the system development or maintenance process [79391]. (b) The software failure incident in Article 79391 was not related to the operation phase. There was no indication in the article that the failure was caused by the operation or misuse of the system. Instead, the incident was clearly attributed to a design flaw in the battery replacement process, leading to the battery catching fire [79391].
Boundary (Internal/External) within_system (a) The software failure incident involving the KiwiBot catching fire at UC Berkeley campus was primarily within the system. The company, KiwiBot, attributed the issue to a faulty battery replacement, specifically mentioning that a functioning battery was replaced with a defective one during a routine battery replacement, leading to the battery experiencing thermal runaway [79391]. The incident was described as an "exceedingly rare occurrence" caused by human error within the system. Additionally, KiwiBot mentioned that they addressed the issue by installing software to monitor the state of the robots' batteries, indicating an internal response to the internal failure.
Nature (Human/Non-human) non-human_actions, human_actions (a) The software failure incident occurred due to non-human actions, specifically a faulty battery replacement. KiwiBot mentioned that the cause of the fire was determined to be human error, where a functioning battery was replaced with a defective one during a routine battery replacement. This led to the rare occurrence of the battery experiencing thermal runaway, resulting in the robot catching fire [79391]. (b) The software failure incident also involved human actions as a contributing factor. The faulty battery replacement, which was the non-human action leading to the failure, was a result of human error during the routine maintenance process. The company acknowledged that the incident was caused by human error in replacing the battery with a defective one [79391].
Dimension (Hardware/Software) hardware (a) The software failure incident in Article 79391 occurred due to hardware issues. Specifically, the incident of the KiwiBot robot catching fire was blamed on a faulty battery replacement. The company, KiwiBot, stated that the cause of the fire was determined to be human error, where a functioning battery was replaced with a defective one during a routine battery replacement. This hardware issue led to the battery experiencing thermal runaway, resulting in the fire [79391].
Objective (Malicious/Non-malicious) non-malicious (a) The software failure incident in Article 79391 was non-malicious. The incident was attributed to human error, specifically a faulty battery replacement that led to the battery experiencing thermal runaway. KiwiBot mentioned that a functioning battery was mistakenly replaced with a defective one during a routine battery replacement, causing the fire in the robot. The company addressed the issue by installing software to monitor the state of the robots' batteries, indicating a non-malicious cause of the failure [79391].
Intent (Poor/Accidental Decisions) poor_decisions, accidental_decisions (a) The software failure incident involving the KiwiBot catching fire at UC Berkeley was primarily attributed to human error, specifically a faulty battery replacement. The incident was caused by a functioning battery being replaced with a defective one during a routine maintenance procedure, leading to the battery experiencing thermal runaway. This indicates that the failure was a result of poor decisions or actions made during the maintenance process [79391]. (b) The software failure incident can also be categorized under accidental decisions as the faulty battery replacement that led to the robot catching fire was described as an exceedingly rare occurrence. It was not an intentional act but rather a mistake made during the maintenance process, highlighting that the failure was due to unintended decisions or actions [79391].
Capability (Incompetence/Accidental) accidental (a) The software failure incident in Article 79391 was not directly related to development incompetence. The incident was attributed to human error during a routine battery replacement, where a functioning battery was mistakenly replaced with a defective one, leading to the battery experiencing thermal runaway. The company, KiwiBot, acknowledged this as a rare occurrence and addressed the issue by installing software to monitor the state of the robots' batteries. (b) The software failure incident in Article 79391 was categorized as accidental. The incident of the KiwiBot robot catching fire was described as an exceedingly rare occurrence resulting from the accidental replacement of a functioning battery with a defective one during a routine maintenance task. KiwiBot stated that the cause of the fire was determined to be human error, indicating that the incident was accidental rather than intentional.
Duration temporary The software failure incident involving the KiwiBot catching fire at UC Berkeley was temporary. The incident was attributed to a faulty battery replacement, specifically a functioning battery being replaced with a defective one during routine maintenance. KiwiBot mentioned that this human error caused the battery to experience thermal runaway, leading to the fire. As a response, the company addressed the issue by installing software to monitor the state of the robots' batteries, indicating that the failure was due to specific circumstances related to the battery replacement process [79391].
Behaviour other (a) crash: The software failure incident in Article 79391 did not involve a crash. The robot caught fire due to a faulty battery replacement, leading to a physical malfunction rather than a system crash [79391]. (b) omission: The software failure incident in Article 79391 did not involve omission. The failure was not due to the system omitting to perform its intended functions but rather a physical issue with the battery replacement [79391]. (c) timing: The software failure incident in Article 79391 did not involve timing issues. The incident was not related to the system performing its intended functions too late or too early but rather a result of a faulty battery replacement leading to a fire [79391]. (d) value: The software failure incident in Article 79391 did not involve value issues. The failure was not due to the system performing its intended functions incorrectly but rather a physical malfunction caused by a defective battery replacement [79391]. (e) byzantine: The software failure incident in Article 79391 did not involve byzantine behavior. The incident was not characterized by the system behaving erroneously with inconsistent responses and interactions but rather a straightforward issue of a faulty battery replacement leading to a fire [79391]. (f) other: The behavior of the software failure incident in Article 79391 can be categorized as a physical malfunction due to a faulty battery replacement, resulting in a fire in the robot. This behavior does not fit into the typical software failure categories but rather falls under a hardware or physical failure [79391].

IoT System Layer

Layer Option Rationale
Perception None None
Communication None None
Application None None

Other Details

Category Option Rationale
Consequence non-human The consequence of the software failure incident described in the article was (f) non-human. The incident involved a KiwiBot robot catching fire due to a faulty battery replacement, resulting in the robot experiencing thermal runaway. This non-human entity, the robot, was directly impacted by the software failure incident [79391].
Domain information, transportation, health (a) The failed system was intended to support the information industry as it was a KiwiBot robot used to deliver food around the UC Berkeley campus [79391].

Sources

Back to List