I'd answer that question by saying that you're absolutely right when you talk about a sequence of events going from research and development to implementation.
The best way to implement technologies is to be working on the right technologies to begin with. A poorly tuned and poorly structured innovation system creates many technologies or products or services and then tries to find an application for them. I must admit, we have hundreds of people or organizations coming to us as COSIA, trying, and this is a figure of speech, to sell the companies a hammer. They're saying, “Could we test your hammer?” when frankly what the companies need is a certain type of screwdriver.
I think a properly tuned innovation system starts with the end in mind. It is a directed innovation: what specific types of technologies do we need, and how would we develop them in ways that allow their implementation to be easy? It's difficult, but we spend a lot of time planning that, but then it's set up so that when people start to work on earlier stages of innovation—ideation, discovery, development, demonstration—they actually don't have to work at implementation. There are companies pulling it; you have companies demanding those types of products.
One thing I would say is that for the government to be part of, first of all, a directed innovation ecosystem that does not just rely on bottom-up good ideas but relies on very clear contact with industry about the commercial application before we start working.... To the extent that the government can be part of that fine-tuned innovation system, it would be very welcome.
There are good examples, but I believe there are opportunities for improvement in this area.