Year of Autonomy
Posted on 11/06/2018
Ian Stock – Knowledge Transfer Manager Rail and Marine
Autonomy and Artificial Intelligence (AI) are now everyday phrases. In 2016 this was not the case when a working group was set up by the South Coast Marine Cluster to look at autonomy not just in maritime, but also across other transport modes.
Tessa Darley, Head of Transport at KTN was asked to chair the group, which consisted of Universities (Portsmouth, Southampton, Plymouth) industry (QinetiQ, Callen Lenz, and later ASV Global) the NOC and Hampshire County Council. Other organisations were involved in the organising of specific events, eg the Plymouth workshop included Plymouth City Council and Plymouth Marine Laboratory
The grouping had the vision to bring together innovators and users from across marine, land and air autonomy to address key challenges and share learning. With the aim of developing cross-sector collaboration to address these challenges.
A programme of 5 workshop events was constructed, which took place over a 15 month period to end March 2018. The topics were as follows:
· Collaborative Autonomy, held at QinetiQ
· Regulation, held at Southampton University
· Risk & Reliability, held at Portsmouth University
· Energy Management, held at Plymouth Science Park
· Societal Acceptance, held at NOC
A short presentation on the progress was made in November at NOC’s MATS Showcase event.
Summary reports and further details are available here
A number of key areas were discussed across the workshops. The three main ones were:
Trust & reliability – societal acceptance
This topic emerged at a number of the events and noted a number of areas for concern and how that concern can be mitigated:
· Certification and regulation set the boundaries within which autonomous systems can operate. Codes of practice can be useful in this respect
· Testing and simulation. How has the system been tested, for how long, and have scenarios been played through sufficiently to give assurance. Simulation allows more scenarios to be tested and a wide range of variations of a particular scenario to be played out.
· Liability, insurance and assurance of safety. Who is responsible for the operational safety, and who takes that responsibility for any problems? Risk management together with codes of practice and certification all contribute to this.
· Evidence and data around simulation and operational use.
· Openness, transparency and auditability around all aspects of production and operations, including integration with human activities and ‘accidents’. Honesty was used in this context.
· Effective communications around current use. A balance between negative and positive reporting of autonomy and AI operations and developments. At present it was seen as too negative.
· A programme of continuous improvement by developers based on experience and the outcomes of in use operations.
Technically a number of challenges remain around:
· Communications both of data collected and emergent issues. This needs development as to how much and what and when can / should be gathered and communicated? It is important when gathered data is not communicated that operations know the system is working / working properly.
· Resilience and endurance, eg related to power and energy, in relation to operational parameters
· Launch and retrieval to and from the water
· Collaboration and connectivity between autonomous systems and with their environment
· Cyber security. This should be designed into the system – to recognize threats. With AI what safeguards can be put in place to make any actions in response to threats proportionate?
Skills, education and jobs
Skills and training need to be appropriate for operations, and integrated into the education and training agenda.
· What is the education system, especially at secondary and tertiary level offering in the way of relevant courses?
· What is the likely impact on existing jobs and roles, and the skills of those already in work? What support mechanisms exist and how can training, reskilling be improved and a clear career path outlined?
The working group will be considering further actions at a future meeting. Summary reports and further information on the year are available here.
However, a recently started initiative is looking at the development, testing and proving the concept of an Autonomous Marine Transport Network (AMTN) on Southampton Water and in the Solent area. Interest in this concept has come from a range of business, research, developer, Government, and other organisations.