Ex-DARPA chief: ‘Innovation with no goals is not efficient’

Ken Gabriel [EIT]

This article is part of our special report Innovation and entrepreneurship.

Ken Gabriel has dedicated his life to innovation and entrepreneurship at Google, and the Defence Advanced Research Projects Agency (DARPA). For Gabriel innovation boils down to “satisfying an unmet need”.

Ken Gabriel is president and CEO of the Charles Stark Draper Laboratory, an independent not-for profit research institution. Before that, he was director of DARPA and deputy director of the Advanced Technology and Projects (ATAP) at Google. He was the keynote speaker at the Innoveit 2016 conference, hosted by the European Institute of Innovation and Technology (EIT) in Budapest.

Gabriel spoke with euractiv.com’s Jorge Valero.

What were the technologies that came out of Darpa while you were there?

I was in DARPA twice. My first period was in the early 90s, when I was recruited as program manager to start the micro-electromechanical systems (MEMS) program. Then I became director of one of the six offices inside DARPA. After that, I left the agency for almost 12 years. I started my own company, that I later sold to Bosch. I was asked to come back to DARPA, first as deputy director and later as acting director.

The things we focussed on during my tenure in the agency was cybersecurity, biotech and biomedicine and hypersonic flights. In regards to personalised medicine, the issue was that we can now read our DNI. That is great but, what do we do with it?

Part of the things we launched while I was there was not only how we analyse the DNA, but how we synthesise from that knowledge. Maybe there is a particular drug or treatment that is good only for you. How do I deliver that in a cost-effective way? Most of pharmaceuticals do not think about individuals in that sense. That fundamentally is an engineering problem from my point of view, not necessarily a biology problem. We also worked in the field of hypersonic flights, flying at Mach 20 (high hypersonic speed), and the challenges associated with it.

What is the secret of DARPA? It is well-known for being a very small team of less than 300 people, with a budget of around $ 3 billion.

It is really about being mission-focused. Mach 20 is a good example. DARPA did not start a hypersonic technology program. It started a Mach 20 program. Going at Mach 20 is a capability. There is no argument about that. If all we did was to develop hypersonic technology  you would say: “Yes, we did some progress in this field.” But did I make enough progress to fly at Mach 20? That is a different focus, and that kind of focus is really important.

The other is the organisational structure. The reason why DARPA is small is because it needs to move with speed and agility. For both things to happen, DARPA needs independence. To me, these are the key characteristics: you need to have independence, a structure that allows for speed and agility and you have to go after bold capabilities, not technologies.

When you compare DARPA with your work at Google, what were the main differences, and what were the results?

Regina Dugan (former DARPA director) and myself were recruited to Google to start the Advanced Technology and Projects group (ATAP). It was different from Google X. The fundamental difference is discipline innovation. What I mean by this is innovation that is disciplined in its execution. There are two schools of thought. I am not saying one is better or worse, but obviously I believe that one is more productive than the other. Google X is more like ‘have great thoughts with no particular objective’, or just ‘do cool things.’ That is the typical approach most people take towards innovation. Give smart people loads of money, a lot of time and good things will come out of it. Actually I don’t think that is right. You do get good ideas coming from that kind of activity, don’t get me wrong. But it is not particularly efficient, nor do you get as much as you could from these kind of investment and people.

The organisation we set up was more like the DARPA model. At Google ATAP, if we hire you in as a technical project leader, we will tell you: you have two years, and after that you have to leave. Culturally that was very hard for people, and also for Google to accept. They would say: why are you forcing these people to leave? To us, that is a very important part of what works at DARPA and what we have seen elsewhere: the focus on execution and discipline. If you join an organization where you have only two years to get something done, you are very focused. You don’t care about your title or the size of your office, or whether you have a parking spot. You are just focused on getting the work done in two years.

There is a lot of talk about a fourth industrial revolution. In your view, what is the most promising technology, or even something that we haven’t seen yet and we can expect?

It would be difficult for me to answer that, not because I don’t understand the technologies, because we are talking about them as technologies. I have yet to have one of these technologies   telling me what unmet need they are satisfying. Only then I would have a better idea of why I would want them. In the case of driverless cars, I can see why they could be useful in terms of congestion and safety. Right now human error is the principal factor for deaths on the road. In the case of virtual reality, or any other technology, until someone tells me what fundamental need is being solved, I will not be able to say which one is more promising.

The problem with many of the technologies is that they talk about them as if the technology itself is the reason why you should get it. That rarely works. People, governments, companies acquire technologies because they give them some advantage, they satisfy some unmet need.

The next generation of mobile internet (5G) will be the technology powering this more advanced digitised world. It seems that the US is less involved in developing the 5G network, compared to Europe or Asia. Why?

I don’t know any specific ratios or numbers, but I do know there is an increasing attention and demand for bandwidth that 5G is going to solve. This is another example of an unmet need that is going to be solved with innovation. There are many US companies and players working on that. But we are in a connected world, so it is not about what we are doing in the US or what Europe is doing. The result is going to be a collection of things going on.

What specific advice would you have for the EIT?

I don’t know a lot about how the EIT is structured or governed. But I do believe that you need to keep organisations that work with innovation relatively small, make as few rules as possible and not try to ‘pick the winners’. There is a tendency to say ‘we have to get it right’ or ‘we cannot put money in companies that fail’. People said about DARPA that we were the organisation that encouraged people to fail. That is wrong. No one should fail. What we do is to encourage people not to fear failure. It is a very different thing. That applies to organisations aswell. If DARPA or Google ATAP took on projects that had to be a guaranteed success, then you pull your risk horizon in. That is not necessarily good for an organisation dealing with innovation.