Written evidence submitted by RACE, UK Atomic Energy Authority (ROB0041)

Key messages

1. In order to lead, the UK should establish world class, real, permanent test facilities for robotics. They will provide a forum to discuss the impact of robotics. They will attract and fix global talent and global investment in the UK. This is achievable and fundable. This market is moving at pace so time is of the essence.

Introduction

2. Robotics and AI are component parts of a maturing digital economy where we collect and analyse data in order to inform decisions and conduct the full panoply of human activities.
3. Robots will be used to collect data and/or conduct ‘physical work’.
4. AI will be used to analyse data to inform human decisions or control robots, i.e. ‘mental work’.
5. These technologies separately and together will become pervasive. They will be disruptive. They will have predictable and unimaginable impacts on the global economy. UK can and should play a leading role.
6. The nuclear sector will increasingly rely on robotics, with need driven by hazardous working conditions. The nuclear sector has a long history of using robotics technology for ‘remote operations’ i.e. enabling operators to work at a safe distance.
7. The nuclear industry will adopt robotics technologies in order to increase safety and productivity of operation in nuclear facilities. This includes decommissioning end of life facilities. Robotics has increased productivity in other sectors by as much as 10%. Applied to decommissioning which is >£100billion liability (=market opportunity), this likely productivity gain should justify significant investment. Life extension of existing nuclear assets and the complete lifecycle of new nuclear, both fission and fusion and both small (e.g. SMR) and large reactors, will be impacted by development of robotics and AI.
8. The UKAEA received £10million funding in 2014 from BIS and OxLEP to establish a new centre for ‘Remote Applications in Challenging Environments’ – RACE. The inaugural Director of RACE, Rob Buckingham FREng, was a primary author of the UK RAS Strategy, July 2014. The RACE facility is now open and the team of 80 engineers represents one of the largest robotics groups in the UK.
9. Definitions:
• Robotics = the next generation of smarter tools (i.e. software controlled hardware such as autonomous vehicles, nuclear robots etc.)
• AI = software that is able to evaluate input data in order to make decisions. Narrow AI is context specific – i.e. algorithms that deal with a particular subset of problems - e.g. autonomous vehicles on a road. General AI represents a much more challenging and more futuristic aspiration of being able to make sense of a much wider set of situations (i.e. more similar to human intelligence). Both narrow and general AI involve machine learning algorithms.
• We will program robots using AI algorithms to improve performance.

The implications of robotics and artificial intelligence on the future UK workforce and job market, and the Government’s preparation for the shift in the UK skills base and training that this may require.

10. The robotics market can be segmented into ‘hazardous environments’ and ‘other environments’. In hazardous or challenging environments (nuclear, space, petrochemicals, mining etc.) robotics solutions are needed to overcome the hazard – i.e. conditions are unsuitable for direct manual intervention.
11. In ‘other environments’ (hospitals, farms, roads, cities) robotics often have safety benefits but the drivers are more likely to be productivity and creating new opportunities (potentially disruptive change).
12. In hazardous environments robotics enable increased economic activity and will create jobs. Robotics will enable humanity to explore and ‘overcome’ the most challenging environments on and off the planet.
13. For example future nuclear power plants will require robotic technology for operation, inspection and maintenance. Robotics will be part of the design of these new plants.
14. We see exactly the same challenge in the petrochemical sector (RACE is working with Shell, Chevron et al. for this reason). Inspection and maintenance in petrochemicals facilities is almost 100% manual (unlike nuclear) but this is likely to change rapidly once robotic solutions are mature. Robotics will allow people working at a safe distance to collect valuable data creating opportunities for data analysis, improved performance and value generation. This value-added activity will create more jobs.

• Key thought: Good data leads to increased knowledge, informs creativity and generates economic activity and jobs.

15. Looking at nuclear decommissioning, the majority of the NDA’s £3billion per annum is spent on wages. More of this could be spent on the wages of innovators (say 3% for 10 years) to create the robotics toolkit that will increase safety and increase decommissioning productivity. If this leads to a 10% saving on >£100billion the ROI is greater than 10x. The types of jobs change and developing decommissioning expertise in the UK will put us in a stronger position to conduct decommissioning around the world.
16. In ‘other environments’, which do not have the same ‘extreme’ hazard, the motivation for use is different, but, notably the technology used is very similar. Robotics is a general purpose enabling technology so it is vital that we invest to enable seamless technology transfer, risk reduction and transfer learning to minimise barriers to market entry.
17. Returning to the jobs issue, in these ‘other environments’ and using autonomous transport solutions as an example, there will be an impact on commercial drivers because the autonomous solutions will be competitive on grounds of safety, price and convenience. In short, effective robotic solutions will release time and money for us to use for other activities and purchases which will create new jobs, but there will be disruption in commercial driving. The speed of transition is critical.
18. This raises two vital issues for national economies. Robotics will be sourced globally. The winners will be those who sell the robotic products and/or services and those who have a flexible, capable workforce able to adapt to changes in employment opportunities.

The extent to which social and economic opportunities provided by emerging autonomous systems and artificial intelligence technologies are being exploited to deliver benefits to the UK.

19. We are in the early stages of a disruptive change. The most advanced robotics systems are autonomous cars. Demonstrations such as Volvo’s DriveMeLondon are vanguard actions: a statement of intent and potential. Widespread product releases are some years away. The most important point to note is that Volvo has come to the UK.
20. Drones have been deployed in Sellafield to map radiation with a human operator ‘in-the-loop’. In other sectors (particularly media) the use of drones is becoming normal but the regulations stipulate line of sight operation. We need to ensure that our regulations enable responsible innovation.
21. In short, there is everything to play for but other countries are spending more than the UK. We risk becoming followers but we can lead if we invest in ‘test sites’ that allow stakeholders to collect real evidence and inform public opinion.

• Key thought: It is early day but the direction of travel is clear.

The extent to which the funding, research and innovation landscape facilitates the UK maintaining a position at the forefront of these technologies, and what measures the Government should take to assist further in these areas.

22. We have a robotics and AI community in the UK which is comparable with other countries. We have to nurture this community and ensure it grows in the UK.
23. The question of how the Government should act to place the UK at the heart of the action has received considerable thought which was consolidated in the UK RAS strategy and other documents. Two specific suggestions are described below – one specific and one generic:

24. CCAV (DfT/BIS Centre for Connected and Autonomous Vehicles) is sending a very strong signal globally that the UK Government is engaged in shaping regulation to allow autonomous vehicles to be tested in the UK first. It is vital to recognise that “autonomous cars are the first robots” and we need to ensure that the UK leads in understanding and shaping this emerging market. It will inform many other robotics markets and public opinion.

PROPOSAL 1: Reinforce and build on CCAV to maximise impact and share learning across Government (health, defence, energy…)

25. The UK should invest in world class, real permanent test facilities. Robots operate in physical environments: e.g. roads, nuclear power plants, hospitals, refineries. These environments have generic requirements (e.g. the need to localise and map) and specific challenges (e.g. pedestrians, radiation, patients and flammable products). Value can only be secured when complete solutions are available in the market. This means combining components to make systems and developing the operating principles and methods to deliver solutions.

26. Reasons for investing to establish world class test facilities include:
27. Showing and demonstrating solutions in a realistic environment enables a conversation between stakeholders: regulators, insurers, users, investors, component and solution suppliers, and the ESPECIALLY the public.
28. Long term demonstration and test generates data that supports evidence based decision-making and investment.
29. Demonstration builds trust. This is vital for both public and private engagement and investment. We need to expresses our concerns and focus on real rather than hypothetical issues.
30. Testing in a real environment is essential for establishing and addressing weakness in the proposition which may be technical, commercial or social. Laboratory testing is not enough.
31. Open demonstration builds understanding at all levels: schools, universities, industry, government. We need to explain the opportunity and the solutions – so that we remain an advanced technological economy ready to use the latest technologies.
32. Test facilities support skills training. Hands on exposure to robotics will encourage the next generation of entrepreneurs.
33. Test facilities attract global interest – e.g. the best academics and multinational R&D dollars will be drawn to facilities that generate activity (invention, start-ups, new products and services…)
34. Test facilities are geographically located. Test facilities act as nucleation sites for clusters, particularly if established close to global brands (cities, universities…).
35. Once established they are difficult to move. Long term facilities are hugely important to anchor a technology in a country and then attract and fix the more mobile aspects: software, finance, insurance, training, regulation.
36. Others are investing. For instance, after detailed consultation led by UK robotics experts, the EU H2020 program has initiated a pilot call for Permanent Test Facilities. In a pilot one consortium will be given €10million to establish a facility and operate a ‘cascade’ program – i.e. funding 3rd parties to use the facility. One such consortium is the SPRINT Robotics Collaborative; a European initiative led by Shell, Chevron, Statoil et al. SPRINT is developing plans for 3 permanent test facilities globally: one in Singapore; one in Houston and one in Europe. Securing such a facility in the UK is vital and should be funded. The addressable market for this one example is estimated to be £750million per annum.
37. Investing in test facilities is a cost effective, achievable ambition. The UK can develop these assets to encourage this technology to stick and stay in the UK.

PROPOSAL 2: Invest now to create word class real test facilities which have open access for developers and end users. Use existing, accessible Government owned infrastructure and facilities to enable maximum engagement and nucleate clusters.

The social, legal and ethical issues raised by developments in robotics and artificial intelligence technologies, and how they should be addressed.

38. Robotics and AI raise fundamental questions about what it is to be human and the direction that society is taking.
39. Research suggests that robotics and AI will transform some employment markets – certain types of jobs will decrease and other jobs will be created.
40. Economists suggest that by creating capital assets (tools) that replace human resources we create the possibility for increased wealth, with time and money being used more productively, but perhaps wealth will become further concentrated.
41. Lawyers raise concerns about liability, but in depth discussion reveals that these are robotics introduces variations on established risks and can be absorbed into existing practices.
42. We wonder whether driverless cars are safe or safer than current technology.
43. Is it ethically and morally acceptable to love your robot? And what does that mean for loneliness in old age and what are the impacts on health care costs?
44. The social, legal and ethical impacts are vast…

• Key thought: Robotics and AI will disrupt employment patterns and change society. Doing nothing is not an option. The UK as a trusted democracy needs to play a leading role.

• “The best way to predict the future is to create it” Abraham Lincoln

• PROPOSAL 3: Establish a high level forum to monitor, pre-empt and respond to social, legal and ethical issues.

April 2016