Advanced Research and Invention Agency Bill – in a Public Bill Committee at 4:10 pm on 14th April 2021.
Q We will now hear oral evidence from David Cleevely CBE, chair of Focal Point Positioning Ltd and the Cambridge Science Centre, and Dr Bob Sorrell, chair of the Campaign for Science and Engineering. We have half an hour for discussion. Could the witnesses please introduce themselves? Thank you very much; it is lovely to see you both.
I am David Cleevely. I am a serial entrepreneur with a background in telecoms and biotech. I have done a lot of work on Government policy, been a board member of the Ministry of Defence and founded networking organisations, including the Centre for Science and Policy at the University of Cambridge. I am glad to see that Bob, who helped me get that off the ground in the early days, is on the panel with me. I am currently chair of the Enterprise Committee at the Royal Academy of Engineering and, for my sins, I am chair of the Cambridge Autonomous Metro Technical Advisory Committee.
Good afternoon and thank you for inviting me to the Committee. My name is Bob Sorrell. I am the chair of the Campaign for Science and Engineering, which is the UK’s leading independent advocate for science and engineering. I come into that with quite a lot of experience in research and development, and I also served two terms on the board of Innovate UK, as well as being a non-executive director in a start-up company, so I have a variety of experience.
Thank you very much. I will start the questions with Chi Onwurah, the shadow Minister for the Opposition.
Q Hi, David and Bob. It is great to see you both and thanks for giving your time. I do not know whether you have been able to follow our evidence sessions, but we have had a lot of discussion about the purpose and remit of ARIA and the extent to which it addresses fundamental failings—if there are fundamental failings—in how the UK chooses and commercialises research. I would like to ask you both your opinion on the remit and purpose of ARIA. What is the failing that you see ARIA fixing? How does it need to change to fix that? Are there things that it is not addressing that it needs to? I will start with David Cleevely.
Thanks, Chi. I would like to start by saying three things rather briefly. First, serendipity does not happen by accident, so we need systems and processes to enable the network diversity and uncovering the unexpected. I am hoping that the new agency will do all of that.
To begin to address some of your other points, we need to improve the whole of the national innovation system. That means not putting in late stage R&D, translation and, in particular—this is something of a bugbear of mine—procurement. If you do not have revenue and if you cannot get product into market, no amount of R&D at the front end will necessarily get you anywhere. If we do not do that, we are always going to be trapped into saying that we need more and more R&D and simultaneously mourning our inability to translate this into economic growth and productivity.
I have one other thing to say, which is slightly cheeky, but I have been listening to the proceedings so far, and they are extremely interesting—it is one of the most interesting sessions I have ever attended. All the examples given of contributions that make a difference have all been, it strikes me, about engineering, so I suggest that we rename this the “Advanced Research and Engineering Agency”. To be honest, “invention” strikes me a bit like something in the 1950s, with somebody emerging from a shed with a gadget that has just blown their hair off. Peter Highnam pointed out “projects”, so we might actually consider it to be the “Advanced Research and Engineering Projects Agency”. No doubt we will get on to why I might say that. The point is that we need to think about this, as Felicity said, in a coherent way, including all the way through to procurement.
Q As a chartered engineer, I am always happy to put engineering in anything. The Advanced Research and Engineering Projects Agency would be AREPA. Your point, which has come up a number of times, is whether this is about, if you like, cutting-edge research, or whether it needs to be looking at transformational translation of existing research, or whether it needs to do both. Certainly, the economist Mariana Mazzucato, as you probably heard, made the point that having the basis there is important, and you seem also to be saying that it needs to look at both, and that it needs to get its purpose right. Let me go now to Bob, and then we will come back.
Thank you very much. Picking up on David’s comments and your question, I am very excited about the potential creation of ARIA. Having something that can respond to the types of challenge that we face, which quite frankly do not respect sector or skills boundaries, is really important today. In particular, there are real opportunities to learn off the back of the covid experience, which has allowed us to really accelerate innovation at quite an incredible pace. If we can take some of that and operationalise that within an ARIA-type environment, that would be a very positive thing.
One thing I have heard, because I have also been listening to the sessions through the day, is mention of crossing the valley of death. For me, there needs to be a matching market pull for the wonderful research products that will come out of ARIA. To get that in place would mean having a really good dialogue between academia and industry and all parties involved to understand what those challenges really are. I also suggest, looking at the DARPA model, that we should back this up by having a really strong public procurement model. Again, we have seen that in covid, and we could see it here, providing a first customer and enabling some of these technologies to be developed. That would be really key.
The final thing I will say is about the personnel involved in this, because that has also come up several times. They really need to have autonomy; they need the ability to make the decisions and choices on what projects they pursue. Equally, they need to be able to start and, critically, stop things. I have much more to say, but I will stop there.
Q Thank you for both of those responses. To follow up, you both emphasised the importance of procurement and market pull. It seems to me that there is such a vast range of areas, issues or challenges that ARIA could look at. What are your views on who should decide what it looks at? Is there a need for a mission, and who should that be set by? If it is left to four, five or six individuals to set the missions, how do we ensure that it is not simply about vanity or pet projects, and that cronyism—we are having some challenges with that at the moment—is not promoted? How do we avoid cronyism and support diversity?
I notice that this came up in the previous session. I think the answer is, in one sense, very straightforward. I think it is for the Government to set the priorities where they feel that there are specific challenges. We have talked about climate change, for example. That is one, and there may be others that one would want to address, either in health or in other topics. That is the point at which the handover occurs and whoever is running ARIA takes that particular domain or challenge. I have been involved, for example, in the Longitude prizegiving, and it was very interesting how we focused down on antimicrobial resistance and testing. A lot of interesting things came out of that. By the way, all the solutions were engineering.
The point is that we should listen to Peter Highnam’s testimony really carefully. Honestly, that was one of the most interesting insights into DARPA that I have had. He talked about the way in which there is autonomy within DARPA to do things within a general area set by Government. Then, within that, there is a peer-review system that enables us to overcome some of the cronyism that you talk about. The more open you are about what you are doing, the less easy it is to hide the fact that you have let particular contracts and so on, so there ought to be a mechanism within the governance structure of the agency to do that.
There is a two-level thing here, but it is up to the Government to decide where the UK’s priorities are. Are we, for example, really concerned about climate change? Can we specify challenges within climate change that will make a difference? In the same way, for defence it was to not be surprised by innovation and to make sure the technology was available for defence in the United States. Within that, DARPA went ahead and looked for things that met that overall goal.
I think there needs to be an overarching structure set for the areas ARIA pursues. In identifying these grand challenges, there is a list that we could reel off right now that would fit the scope. Earlier, I heard conversations about having six wise people who would make these decisions and cover these areas. I worry about that approach. I think you need people who are really up for engaging people to understand the nature of the problems and translating them into meaningful challenges.
The other part that is often missed in this is the social science aspect, because there has to be a level of public acceptance around the things that people are developing on their so-called behalf, and that part is also incredibly important. We need to have a very open process for how we decide on those projects so that we avoid, as you say, falling into the traps of vanity or pet projects. If you have clear criteria from the outset and stick to them, you will be fine in that regard.
Thank you. I could ask you lots more questions, but I will stop there and hand over.
Q It is lovely to see our two witnesses this afternoon. The first question is to Bob. You mentioned in your written evidence to the Science and Technology Committee that a body like ARIA will need an understanding of failure. What does a new understanding of failure look like, exactly?
That is a great question. If you compare and contrast us with the Americans, there is a definite culture in the UK that failure is something that you hide under the carpet, put away and forget, but science is all about failure and pushing the boundaries. If you are not failing, you are really not challenging those boundaries. I think it is about establishing a culture in which we can accept failure and move on.
The problem comes, in both industrial and academic environments, in facing that day, because there is a tendency to keep things creeping along because you have invested so much effort to get them to this particular point. You do not want to kill it, because then you have to stop the project, and people feel emotionally involved in it. That creates a whole series of issues associated with it. It is about making the hard decisions and learning from failures. We describe them as failures, but actually they are some of the most valuable learning experiences that we gain, and they stop us reinvesting in making the same mistakes in the same areas if we are really careful about what we extract from them, and do not just try to shut them off in a box, in a rather embarrassed way, and say, “That’s something that we will leave to one side.”
Q Thank you very much for that. I have a more general question to both of you. One of the things we have heard a lot about today is the importance of fitting ARIA into a larger innovation system. What advice do you both have about how ARIA can forge the most productive relationships with a range of public and private sectors, for example?
As well you know, I am very keen on establishing networks of individuals and making sure there is lots of exchange. Part of the essence of putting an agency like this together is to ensure that you get a lot of cross-fertilisation. There should be a great deal of exchange going with that, and you would, of course, have to have in place the conflict of interest and various other peer-review processes.
It is very important that an agency like this would work closely with the private sector. My first encounter with DARPA goes back to 1977. At that point, I was working for Post Office Telecommunications, which shows how long ago it was. We were discussing the idea of funding this funny thing where you cut information up into packets. A lot of the collaboration that was done on all of that involved a great deal of what was then a monopoly, though a commercial entity, helping to fund those things. That kind of stuff is extremely important and needs to be built into the processes by which this agency operates.
Can I just pick up on the notion of failure? There are two kinds of failure. There is the kind of failure that we have seen with SpaceX, where you send a rocket up and you land it and it crashes or burns up after about 12 minutes because it is leaking fuel. That is one kind of failure. Quite honestly, the private sector got involved in replacing NASA because NASA became too cautious about dealing with that kind of thing.
There is another kind of failure where you have picked the wrong technology—the wrong way of approaching a problem. I think we are talking about the second kind, and about recognising how to stop that. That is a peer-review process; that is a way of making sure you do things. What we need to avoid is reacting to failure where the rocket is crashing on touchdown. That is not really failure; that is simply experimentation.
Q Thank you, David. Bob, any further thoughts on the relationship and how to engage with public and private sector?
I would say three quick things. First, ensure that there is a real partnership between industry, Government and academia, in actually shaping the agenda for ARIA. I would have flexibility; we heard that earlier, I think, from a colleague from the CBI talking about models in which we could second people into the ARIA organisation. I think there is an opportunity to do that, and we have had experience of doing that previously.
The other thing is that ARIA provides some really important learnings, and it should be able to integrate those back into UKRI, and vice versa. UKRI has some valuable learnings that it can impart to ARIA. This is an evolutionary process through which both parties will definitely benefit, and it should be framed in that light.
Q Thank you, Chair. Welcome, both of you; it is good to see you. I have been struck throughout today by the evidence that we have been looking at an American model, trying to learn lessons from it and importing it into what looks to me like a rather different environment and landscape here. I would like both of you to ignore all that we have heard today and answer the question I would ask at the beginning, which is: what is the big problem facing the UK R&D system? Do you think ARIA, as it is presented, can be bent to fit whatever problem it is that you think needs to be solved?
What is the big problem? The big problem is that we do not have procurement systems that buy enough stuff from small and medium-sized enterprises. Half the employment growth in this country comes from 7% of the SMEs that are fast growing. If you look at a place like Cambridge, as you well know, Daniel, we have 20 $1 billion companies. Companies that have come into existence that were not even a glimmer in somebody’s eye in 2014 and are now about to be floated.
That is the kind of process we need to understand, and why we do not have more of those successes. In particular, if I may blow Cambridge’s trumpet, we need to understand why we have those things happening in Cambridge, and why they are not being replicated elsewhere. From my personal point of view, having sold a company to an American buyer last November, which, as you can imagine, was an interesting experience, it was because it had innovative technology. We were absolutely the best in the world and hardly anybody from the UK bought anything from us. The majority was being bought by Americans—American defence and security stuff.
It is a great disappointment to me that we do not have the ability to nurture and bring on. The way the Americans do it is that they have that complete system. They have an awful lot of money and effort going into procurement. Somebody spoke earlier about the infantilisation of Government Departments, and the way in which that expertise is not there. I will mention engineering again here. We need more engineers in Government, who can take those kinds of decisions and understand what we need to procure to be able to do things. That strikes me as so important. It is not to detract from AREPA, as we might call it, but in order for it to be as functional and effective as possible, we need to look at the entire system.
That was a great answer from David. I will just pick up on a couple of things. I go back to the fundamental issue of matching the research that is coming out of ARIA with the market pull for it. It is important to define what the challenges are up front. The role of public procurement, as David raised, is critical, as is supporting the growth of the so-called Mittelstand—the mid-sized companies—and understanding what is behind the culture that leads to so many of those companies being sold at around the £50 million level, as opposed to growing to the hundreds-of-millions-of-pounds companies that they could be. How do we support them through that whole growth cycle? There is much more that I think we could do in that space.
Q To summarise slightly—I do not want to become a grim AREPA—what we need is not only an ARIA that can do the things that we have been talking about today, but possibly other things alongside it to make it work. Would you agree with that?
It is fine tinkering around with the engine and putting another turbocharger on it, but if the chassis, the transmission system and the wheels will not deliver what you need, all that energy and power is going to go somewhere. In an international system, all we will do is to help to accelerate other countries that are willing to buy our stuff from us. That is fine; I am all for international co-operation, but I really would like to see a bigger contribution to economic growth and productivity improvements in the UK.
To pick up on what David is saying, ARIA is part of the solution. We need all the things that we have, effectively, to put us in a position to lead against the challenges that we face. We would not be in this position if we did not have such a brilliant research community in the UK to start with. It is fantastic that we are having a conversation about how we capitalise on that. It is not just £800 million for ARIA, which is just seed money to start it, but the investment in the overall infrastructure that will make many of these things possible. We need to commit to doing that as well, if the UK is really going to lead and be the test bed and demonstration centre for the technologies that it can lead in and deploy globally.
Q If nobody else has a question, I will take the opportunity to come back on that point, particularly on procurement. I remember having a great deal of difficulty persuading the Prime Minister that the American Department of Defence was far better at buying stuff from UK small businesses than the UK Ministry of Defence, as the figures show. What do you think we could do, or what should Government be doing, to enable, require or ensure that ARIA, or AREPA, better supports small business growth and, at the same time, addresses the issue of market pull?
The general thrust of what AREPA—if we are going to adopt that word—is trying to do is right. There are a number of things going on in bits of defence, for example. You have DASA and various others playing around with projects within the different services, for acquiring different kinds of technology. I think the phrase “a bit more coherence” was used by Felicity. I think we need to understand what the map of that innovation system looks like.
I am pretty convinced that people are pretty smart—they will make the right decisions. You just need to give them the right structure, hence my point that serendipity does not happen by accident. These kinds of things happen because you have constructed systems and processes so that people bump into and talk to each other, and will exchange ideas. ARIA is fine as it stands, but it sits within quite a complex system. I would like to see much more recognition within Government about how complex that system is, and how it actually operates. I completely agree with you that it has been far easier, in all my companies, to sell stuff into the United States—particularly into the United States defence market—than it has ever been to sell into the UK.
To build on that, I did a couple of terms at Innovate UK and we tried stimulating public procurement during that period. I think a lot of it is about the culture and getting it right, to allow people to invest in those smaller companies and different technical solutions, to move them away from the existing ones. We got that to work during covid. We managed to get it to work, and we managed to get ourselves investing and procuring things in a different way. That is why I keep coming back to that and looking at what we did differently then that allowed people to make those different choices. I think we have to take some of that learning to see how we can get public procurement to work in a better way going forward.
Q Bob, you mentioned engagement and trust. We have heard a lot today about accountability and trust. How do you feel that we can get that trust without stifling innovation, and do you think FOIs are the best way to do that?
If you are to get trust, you need to be transparent about the choices that you are making and how you are making them. Then, when you move to the execution phase, you need to allow the programme managers and the people who are driving the programme scenario to make the choices flexibly and in the quickest way possible. I understand in part what you are perhaps playing into, but I think you just need to strike the right balance between transparency on how choices are made and holding to account on that, and allowing people to get on with executing against those programmes once those choices have been made.
I think the acid test is whether you can explain something to someone who is independent and is one of your peers. If you are happy explaining it back to somebody like that, that is fine. That is the way in which the system works. If you listened to Peter Highnam talk about how DARPA was organised, that was built into the DNA.
Q Do you think it would be useful to have that built into ARIA from its inception?
I think it is essential. I would be very uncomfortable if you had an agency that did not have some degree of—accountability is not exactly the way to describe it, but you have to have a group of independent people reviewing what you are doing, not quite in the same way as you would do an audit, but it is basically that kind of principle. If I have to explain something, as I am having to do for this Committee, it is a lot clearer and more straightforward, and I feel a lot more comfortable about the way in which I can rely on the ideas and what I am doing. I think that process is very, very important.
If there are no further questions from Members, I thank the witnesses for their evidence. The Committee will meet again on Tuesday at 9.25 am to begin line-by-line consideration of the Bill.