As the adoption of artificial intelligence (AI) technologies advances digital transformation, AI continues to be top of mind for a variety of stakeholders worldwide, from start-ups to AI experts in large industry, academia, research and regulatory.
The IEC and ISO Joint Committee for artificial intelligence (ISO/IEC JTC 1/SC 42) held its Plenary in Tokyo last month, where over 30 resolutions were passed and new areas of work begun. The event was hosted by the Japanese National Body.
“AI continues to fuel the digital transformation, which will in turn change how we live, work and play for generations to come”, said Wael William Diab, who chairs SC 42. “SC 42 is developing projects and studying emerging areas across the entire ecosystem”.
Standards in the pipeline
Over 50% of the SC 42 programme of work has now moved past the initial working draft state, with the Big Data Reference Architecture moving into final balloting. International standards projects moving to the next stage of development included:
AI ecosystem projects
- ISO/IEC 22989 Concepts and terminology
- ISO/IEC 23053 Framework for AI systems using machine learning
- ISO/IEC 24030 AI Use Cases
- ISO/IEC 24029-1 Assessment of robustness of neural networks Part 1: Overview
Big Data ecosystem projects
- ISO/IEC 20547-1 Big data reference architecture – Part 1: framework and application process
- ISO/IEC 20547-3 Big data reference architecture – Part 3: reference architecture
New areas of study initiated
To address emerging areas of interest and challenges in the broad adoption of AI, SC42 established two new Advisory Groups:
- SC 42/AG 1 AI Management Systems Standard (MSS) is
tasked with preparing a justification study which identifies the elements and
substantiates the viability of a management systems standard. AI technologies present
concerns beyond traditional IT systems such ethical and trustworthy
considerations in the training data used. A potential MSS can allow for
assessment of the AI processes used. The Group will submit its report at the
next SC 42 Plenary in April 2020.
- An AI Systems Engineering Advisory Group was set up to carry out, among other things, a gap analysis between existing engineering practices and AI best practices to identify standardization opportunities. The group will also study system engineering challenges unique to AI systems such as adaptation, integration, lifecycle implications, redeployment and maintenance.
SC 42 covers the entire AI ecosystem with an extensive network of liaisons that include many committees within JTC 1, IEC and ISO. It also collaborates with external groups. During the meeting, liaisons were approved with ITU-T, Partnership on AI (PAI), ISO TC 37/SC 3 Management of terminology and IEC TC 65/SC 65A System aspects of industrial-process measurement, control and automation. SC 42 is inviting all its liaison partners to submit relevant use cases on artificial intelligence.
More about the committee for AI
SC 42 has published three standards, has 13 open projects and six working groups. It is comprised of 29 participating members and 12 observing members. Over 150 delegates and 170 experts attended the Plenary in Tokyo.
Diab thanked the event host on behalf of the committee – Japanese National Body as well as the Ministry of Economy Trade and Industry (METI), Information Processing Society of Japan (IPSJ) and the Information Technology Standards Commission of Japan (ITSCJ). He also acknowledged the meeting sponsors – National institute of advanced industrial science and technology (AIST), Mitsubishi Electric Corporation, NEC Corporation, Hitachi Ltd and Microsoft Japan.