This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
Building Responsibility in AI: Transparent AI for Highly Automated Vehicle Systems
Technical Paper
2021-01-0195
ISSN: 0148-7191, e-ISSN: 2688-3627
Annotation ability available
Sector:
Event:
SAE WCX Digital Summit
Language:
English
Abstract
Replacing a human driver is an extraordinarily complex task. While machine learning (ML) and its’ subset, deep learning (DL) are fueling breakthroughs in everything from consumer mobile applications to image and gesture recognition, significant challenges remain. The majority of artificial intelligence (AI) learning applications, particularly with respect to Highly Automated Vehicles (HAVs) and their ecosystem have remained opaque - genuine “black boxes.” Data is loaded into one side of the ML system and results come out the other, however, there is little to no understanding at how the decision was arrived at.
To make these systems accurate, these AI systems require lots of data to crunch and the sheer computational complexity of building these DL based AI models also slows down the progress in accuracy and the practicality of deploying DL at scale. In addition, the training times and the forensic decision investigation — often measured in days, sometimes weeks and months — slows down implementation and makes traditional agile approaches with their definition of done almost impossible to follow.
Recent breakthroughs have allowed ML systems in a HAV implementation context to determine reasonable solutions in very fixed scenarios. However, these systems are typically very complex and largely incapable of explaining how or why they came up with that solution. Without this knowledge and reasoning, intervention and proof of compliance during HAV development, validation, verification, and production applications is near impossible. To cut development and forensic time it takes to create and understand DL models with high precision, decisions must be understood, and reasoning applied.
While significant breakthroughs have been made in Explainable AI (XAI) through DL technologies such as recursive methods, and Cognitive AI (CAI) through user interfaces (UI), they all commonly fail at “transparency”. Transparency is the ability to have access to the logic behind a decision made by a ML system. This is a requirement to establishing trust in high risk and high human cost applications such as an HAV. This paper will outline how a solution based on Knowledge Representation and Reasoning (KRR) creates a “holistic AI” approach that enables both knowledge on how a HAV machine learning system arrives at decisions, and provides the rational or reasoning through the provisioning of new insights into what would typically be a blind process. This “Transparent AI” solution will be explored through an algorithmic approach and then demonstrated through a software implementation within Baidu’s Apollo model framework.
Authors
Topic
Citation
Minarcin, M., "Building Responsibility in AI: Transparent AI for Highly Automated Vehicle Systems," SAE Technical Paper 2021-01-0195, 2021, https://doi.org/10.4271/2021-01-0195.Also In
References
- National Science and Technology Council 2016 https://www.whitehouse.gov/sites/default/files/whitehouse_files/microsites/ostp/NSTC/preparing_for_the_future_of_ai.pdf
- National Science and Technology Council 2016 https://www.nitrd.gov/PUBS/national_ai_rd_strategic_plan.pdf
- European Parliment Feb. 16, 2017 www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//NONSGML+TA+P8-TA-2017-0051+0+DOC+PDF+V0//EN
- Bo , X. May 23, 2016 http://news.xinhuanet.com/english/2016-05/23/c_135382029.htm
- The Guardian/Reuters Oct. 10, 2018 https://www.theguardian.com/technology/2018/oct/10/amazon-hiring-ai-gender-bias-recruiting-engine
- Newman , D. Your Artificial Intelligence Is Not Bias-Free Forbes Sep. 12, 2017 https://www.forbes.com/sites/danielnewman/2017/09/12/your-artificial-intelligence-is-not-bias-free/#25165446c783
- Mittelstadt , B. , Allo , P. , Taddeo , M. , Wachter , S. , and Floridi , L. The Ethics of Algorithms: Mapping the Debate Big Data & Society Dec. 1, 2016
- Careaga , A. April 10, 2018 https://phys.org/news/2018-04-uber-tesla-incidents-artificial-intelligence.html
- Burgess , M. Jan. 11, 2016 https://www.wired.co.uk/article/creating-transparent-ai-algorithms-machine-learning
- Burt , A. The AI Transparency Paradox Harvard Business Review Dec. 13, 2012 https://hbr.org/2019/12/the-ai-transparency-paradox
- King , D. Putting the Reins on Autonomous Vehicle Liability: Why Horse Accidents Are the Best Common Law Analogy North Carolina Journal of Law & Technology 19 4 126 159 Jan. 01, 2018
- Press , G. Artificial Intelligence (AI) Defined Forbes Aug. 27, 2017 https://www.forbes.com/sites/gilpress/2017/08/27/artificial-intelligence-ai-defined/?sh=6ec1c9f57661
- KPMG 2017 https://home.kpmg.com/content/dam/kpmg/us/pdf/2017/02/I-see-I-think-I-drive.pdf
- Next Move Strategy Consulting Oct. 2020
- Government Technology Nov. 14, 2016 https://www.govtech.com/computing/Understanding-the-Four-Types-of-Artificial-Intelligence.html
- Marenus , M. https://www.simplypsychology.org/multiple-intelligences.html
- Cohen-Yashar , M. https://www.linkedin.com/pulse/expert-systems-knowledge-based-ai-manu-cohen-yashar/
- Brooks , R.A. https://people.csail.mit.edu/brooks/papers/how-to-build.pdf
- Freska , C. Qualitative Spatial Reasoning Cognitive and Linguistic Aspects of Geographic Space 361 372 1991
- Dague , P. Qualitative Reasoning: A Survey of Techniques Applications AI Communications 8 3-4 119 192 July 1995
- Hernández , D. Qualitative Representation of Spatial Knowledge. Lecture Notes in Artificial New York Springer Verlag 1991
- Banavar , G. Nov. 30, 2016 https://hbr.org/2016/11/what-it-will-take-for-us-to-trust-ai
- Selyukh , A. Sep. 28, 2016 https://www.npr.org/sections/alltechconsidered/2016/09/28/495812849/tech-giants-team-up-to-tackle-the-ethics-of-artificial-intelligence
- Gunning , D. Nov. 2017 https://www.darpa.mil/attachments/XAIProgramUpdate.pdf
- Buhrmester , V. , Münch , D. , and Arens , M. Nov. 27, 2019 https://arxiv.org/pdf/1911.12116.pdf
- Knight , W. April 11, 2017 https://www.technologyreview.com/2017/04/11/5113/the-dark-secret-at-the-heart-of-ai/
- Bojarski , M. , Yeres , P. , Choromanaska , A. , Choromanski , K. et al. April 25, 2017 https://arxiv.org/pdf/1704.07911.pdf
- Bennett , T. Oct. 8, 2018
- Bojarski , M. , Jackel , L. , Firner , B. , and Muller , U. May 23, 2017 https://developer.nvidia.com/blog/explaining-deep-learning-self-driving-car/
- Hernandez , D. Qualitative Representation of Spatial Knowledge - Subseries Lecture Notes in Computer Science 804 New York Springer-Verlag 1994
- Hunter , H. Nov. 22, 2017 https://venturebeat.com/2017/11/22/its-time-to-solve-deep-learnings-productivity-problem/
- Dickson , B. July 29, 2020 https://bdtechtalks.com/2020/07/29/self-driving-tesla-car-deep-learning/
- Samek , W. Jan. 09, 2019 http://heatmapping.org/slides/2019_NLDL.pdf
- Marcus , G. July 15, 2017 https://arxiv.org/ftp/arxiv/papers/1801/1801.00631.pdf
- Leigh , J. https://medium.com/method-perspectives/the-evolution-of-trust-ad0b9bfc7e5f
- von Simson , C. April 01, 2019 https://blog.rossintelligence.com/post/transparency-is-the-key-to-ethical-ai-decision-making
- Escrig , T. and Toledo , F. Qualitative Spatial Reasoning: Theory and Practice: Application to Robot Navigation Amsterdam IOS Press 1998
- Kuipers , B. Qualitative Reasoning: Modeling and Simulation with Incomplete Knowledge (Artificial Intelligence) Cambridge, MA The MIT Press 1994
- Stèpánková , O. An Introduction to Qualitative Reasoning Advanced Topics in Artificial Intelligence, Lecture Notes in Artifical Intelligence 617 Prague Springer 1992 404 418
- Minarcin , M.A. , Kim , S.-I. and Escrig , T. April 23, 2018
- Escrig , T. and Chung , S. Lessons Learned from Robotics Applied to Cyber Security International Journal of Computer Applications 74 8 12 18 2013
- Mitsubishi Electric Corporation https://us.mitsubishielectric.com/en/news/releases/global/2020/0128-a/index.html
- Wordpress Sep. 04, 2010 https://artificialintelligentsystems.wordpress.com/tag/knowledge-representation-and-reasoning/
- Werthner , H. Qualitative Reasoning: Modeling and the Generation of Behavior Vienna Springer 1994
- Mahto , M. , Hogan , S.K. , and Hatfield , S. https://www2.deloitte.com/us/en/insights/focus/technology-and-the-future-of-work/machine-learning-qualitative-data.html
- Freksa , C. and Röhrig , R. Dimensions of Qualitative Spatial Reasoning Qualitative Reasoning in Decision Technologies 483 492 1993
- KupperingCole https://www.youtube.com/watch?v=37p6G9WNmNk
- McKenna , B. Aug. 07, 2019 https://www.computerweekly.com/blog/Data-Matters/The-Enterprise-Data-Fabric-an-information-architecture-for-our-times
- Deus , H. Oct. 18, 2018 https://www.linkedin.com/pulse/knowledge-graphs-machine-learning-iswc-2018-trip-report-helena-deus/
- Marr , B. July 04, 2019 https://www.linkedin.com/pulse/knowledge-graphs-machine-learning-future-ai-analytics-bernard-marr/
- Gruber , T. 2009 http://www-ksl.stanford.edu/kst/what-is-an-ontology.html
- OWL Working Group Dec. 11, 2013 https://www.w3.org/OWL/
- National Safety Council https://www.nsc.org/safety-training/defensive-driving/courses/classroom/ddc-10th-edition
- United States Department of Transportation 2020 https://www.nhtsa.gov/road-safety
- Glez-Cabrera , F.J. , Álvarez-Bravo , J.V. , and Díaz , F. QRPC: A New Qualitative Model for Representing Motion Patterns Expert Systems with Applications 4547 4561 Sep. 1, 2013
- Glez-Cabrera , F.J. , Álvarez-Bravo , J.V. , and Díaz , F. Representing Motion Patterns with the Qualitative Rectilinear Projection Calculus Distributed Computing and Artificial Intelligence 251 258 2013
- Álvarez-Bravo , J.V. , Peris-Broch , J. , Escrig-Monferrer , M. , Álvarez-Sánchez , J. , and González-Cabrera , F. A Qualitative Representation Model about Trajectories in 2-D Frontiers in Artificial Intelligence and Applications 146 27 124 May 2006
- Álvarez-Bravo , J. , Peris-Broch , J. , Álvarez-Sánchez , J. , and Escrig-Monferrer , M. 2006
- Baidu June 29, 2019 https://github.com/ApolloAuto/apollo/releases/tag/v5.0.0
- Delmolino , D. and Whitehouse , M. 2018 https://www.accenture.com/_acnmedia/pdf-92/accenture-afs-responsible-ai.pdf
- IBM IBM CEO Ginni Rometty to Open CES 2019 with Keynote Address Business Wire Aug. 1, 2018 https://www.businesswire.com/news/home/20180801005015/en/IBM-CEO-Ginni-Rometty-Open-CES-2019
- Reportlinker.com Aug. 2017 https://www.reportlinker.com/p05090296/Automotive-Artificial-Intelligence-Market-by-Offering-Technology-Process-Application-and-Region-Global-Forecast-to.html