Computer Science > Artificial Intelligence
[Submitted on 20 Jul 2024]
Title:A Measure for Level of Autonomy Based on Observable System Behavior
View PDF HTML (experimental)Abstract:Contemporary artificial intelligence systems are pivotal in enhancing human efficiency and safety across various domains. One such domain is autonomous systems, especially in automotive and defense use cases. Artificial intelligence brings learning and enhanced decision-making to autonomy system goal-oriented behaviors and human independence. However, the lack of clear understanding of autonomy system capabilities hampers human-machine or machine-machine interaction and interdiction. This necessitates varying degrees of human involvement for safety, accountability, and explainability purposes. Yet, measuring the level autonomous capability in an autonomous system presents a challenge. Two scales of measurement exist, yet measuring autonomy presupposes a variety of elements not available in the wild. This is why existing measures for level of autonomy are operationalized only during design or test and evaluation phases. No measure for level of autonomy based on observed system behavior exists at this time. To address this, we outline a potential measure for predicting level of autonomy using observable actions. We also present an algorithm incorporating the proposed measure. The measure and algorithm have significance to researchers and practitioners interested in a method to blind compare autonomous systems at runtime. Defense-based implementations are likewise possible because counter-autonomy depends on robust identification of autonomous systems.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.