From the UK Prime Minister Theresa May at a Downing Street reception held for tech sector representatives last night, a bold statement of intent:
The economy fit for the future which we will build as we leave the European Union, and forge a new role for ourselves in the world, must be one which leads the world in innovation and emerging technologies.
Not mentioned last night was an unfortunately timed report from the National Audit Office (NAO) which stated that, for all the upbeat talk around AI and robotics from various government ministers, up to and including the PM, the UK actually doesn’t have a strategy for investing in this sector, nor is it providing the leadership necessary to establish one.
As part of a wider report into overall goverment R&D programmes - Cross-government funding of research and development - Amyas Morse, Head of the NAO, warns:
Some areas of research have well-established arrangements to support coordination and collaboration between public-sector funders. But some newer areas, including important emerging technologies and areas of national importance, need more effective leadership. As a proportion of GDP, the UK spends less on research and development than many comparable nations. Government needs a coherent view of the UK's research strengths relative to other nations and analysis of funding in key areas of research, so that it can prioritise areas where activity is lagging behind and ensure the UK is investing in the right areas.
The NAO found that since 2012 the government, in one form or another, has spent - or said it plans to spend - over £300 million in robotics and autonomous systems (RAS) research. But it adds the caveat:
Total investment in RAS research is not reported consistently or in one place, which makes it difficult to calculate an accurate figure.
What is known is that government funding to date includes a joint £100 million government/industry programme to support driverless cars in the UK, £100 million of research council grants, £25 million of Innovate UK funding for RAS-related innovation projects, and £32 million for research into unmanned aerial vehicles.
The NAO reports that in 2015, the UK spent £31.6 billion on R&D. That’s less as a percentage of Gross Domestic Product (GDP) than the average for European Union (EU) and Organisation for Economic Co-operation and Development (OECD) countries.
The EU has set a target to increase combined public and private investment in R&D to 3% of GDP by 2020. But of course the UK will, theoretically, have exited the EU by that point, so who knows what will happen.
The NAO also notes that Brexit is likely to bring fresh pressures on R&D as the UK is currently a net receiver of competitive EU funding for research. Between 2007 and 2013, the UK contributed €5.4 billion and received €8.8 billion. For good measure, the NAO, citing the Council for Science and Technology, points out that an estimated 80% of funding for UK RAS research comes from the EU.
But that’s all to come. For now, the point is made that the UK’s investment in robotics and machine-learning just isn’t up to speed, this despite the recognition in the January 2017 green paper on the nation’s Industrial Strategy emphasising the importance of research for economic growth.
A new body, UK Research and Innovation (UKRI), is being set up to bring together research councils, Innovate UK and Research England (HEFCE’s research funding functions) to create an integrated research and innovation system. But this will not be in place until April next year.
Until then, the Department for Business, Energy & Industrial Strategy (BEIS) has responsibility for the majority of government investment in science, funding R&D through partner organisations, esearch councils, Innovate UK and the Higher Education Funding Council for England (HEFCE). Some other departments also fund specific robotics R&D - such as the Ministry of Defence.
And here’s where the problem with the robot revolution arises - BEIS just isn’t doing its job properly. The NAO report says:
Our examination has demonstrated that strong leadership is the driving force for coordination and making everything else happen – this includes setting priorities, and having good information to make decisions and evaluate the impact of investment…Despite consensus that it is needed, BEIS has not yet established government leadership and a strategy for investing in robotics.
What is needed is a change of approach, says the NAO :
Collective action is needed to prioritise investment in three research areas we examined, to ensure efforts are focused on addressing the principal challenges…In robotics and autonomous systems, and advanced materials, key players have identified strategic themes for investment but a top-down strategy would help reach consensus on priorities.
In robotics, research underpins and supports scientific advances in a range of technologies, so a joined-up approach across government to set cross-cutting priorities is important. While funders may have differences in focus, the establishment of a leadership group can provide a focal point for discussion, and improve understanding of others operating in the same space, thus helping to better direct investment.
The impression is given that what exists today is akin to a turf war:
For cross-cutting technologies like robotics, where funders and other stakeholders are still emerging, there are additional challenges in establishing who is involved and what research they are funding. For example, robotics has many applications but no main sponsoring department, making it difficult to track and classify research activity. Funding may only be classified as robotics research at the point when the technology
has become an end product. Stakeholders have told us that the cross-cutting nature of robotics and autonomous systems makes coordination difficult as there is no single responsible department.
This isn’t a new problem, notes the NAO, citing a 2014 strategy document from the Robotics and Autonomous Systems Special Interest Group, funded and managed by Innovate UK’s Knowledge Transfer Network, which pushed for the establishment of a leadership council to bring funders together and engage with leaders across industry, academia and government. This was accepted and approved by BEIS, in its former guise as the Department for Business, Innovation & Skills.
Flash forward three years and the end result is…zilch. And it’s not as though BEIS hasn’t been reminded of the importance of some action here:
In October 2016, the House of Commons Science and Technology Committee flagged up concerns about the continuing lack of leadership and absence of a strategy for developing skills and securing the investment needed for further growth. This brought another set of approving noises from BEIS, but still no actual action:
BEIS has said that funding from the new Industrial Strategy Challenge Fund will be available to support priorities such as robotics, and that it would consider the best model of leadership for robotics as it develops its industrial strategy. However, the department has not yet confirmed the arrangements for ensuring strategic oversight and coordination.
Among the NAO’s recommendations is that the UK Research & Innovation (the new organisation which will be in place from April 2018 to create an integrated research and innovation system) and BEIS should begin work with other government departments and the Government Office for Science to identify the areas of research that need strategic leadership and co-ordination. The NAO has separately published an evaluative framework to help funders assess and develop leadership arrangements in their areas of research.
It’s to be hoped that Greg Clark, Secretary for State at BEIS, found time before the cocktail nibbles last night to take a long, hard look at this report.
At the Downing Street reception, the Prime Minister referenced him as one of the people in government driving through the digital revolution.
Well, Prime Minister, he and the rest of BEIS might be driving, but they’re most definitely in the slow lane and in danger of being lapped by foreign competitors.
As Meg Hillier, chair of the influential House of Commons Public Accounts Committee, said yesterday:
Quite. That lack of leadership is a common complaint across all of pre-Brexit government digital policy-making. Lest I be accused of picking on BEIS, let’s state clearly that there’s precious little sign of much to be commended coming out of the Department for Digital, Culture, Media and Sports, unless you count ministerial Instragram accounts as a big tech moment.
As members of the EU, we have access to European projects, free movement of researchers, and billions of pounds of investment in the UK from the EU. There is a risk we could lose our edge as a research power as a result of Brexit. In order to avoid that we need strong leadership from Government departments and UKRI. But to date, in important areas like robotics and climate change, that leadership has been sorely lacking.
This report is a timely, if depressing, reminder of just how much has to change. When the Indsutrial Strategy is published, there will be lots and lots of inspirational talk about the UK’s track record in technological innovation. The Prime Minister was at it last night, name-dropping Charles Babbage, Ada Lovelace and Alan Turing, the latter feted as "the visionary mathematician who contributed so much to the development of Artificial Intelligence".
Their past achievements are rightly recognised. But I can’t imagine Turing would find the present day commitment to R&D remotely acceptable in terms of developing the nation's technology future. Or had much polite small talk to swap over the Downing Street cocktail sausages last night.