For me, one of the many joys of working at a university is the freedom to think about interesting problems such as this, and to talk to many different people in many different organisations about them. In this case, our discussions slowly converged into a crazy idea: could this problem be solved by using, instead of a single sensor, a swarm of miniature sensors? Over time, and with growing knowledge of their environment, they would be able to provide better and better information.
The only way to find out was to set up a large project to investigate further. It would be inherently risky and costly, and would have to involve international partners, but if successful would be highly beneficial to society: a perfect fit for the EU’s Horizon 2020 FET Open funding programme. A consortium of institutions submitted an application. To our delight, the Phoenix project was approved.
Running this project is every bit as challenging and exciting as you would expect. We aim to make our crazy idea usable and used for many societally relevant applications – addressing primary human needs such as energy, safety, food and water, and health. In the first phase of the project, targets and boundary conditions had to be generic and thus vague. Another challenge was the composition of our team: 13 students with a wide range of specialisations ranging from ultrasonic transducer design to knowledge elicitation. Learning to understand each other and see how our different tasks would together make this idea a reality was a significant challenge. We also needed a non-standard approach to integrating such a diverse team, as described by project member Karine Miras in her blog post “Together is the Matter”.
The bankruptcy of one of our partners during the first half of the project was a major and unexpected disruption. To keep the majority of researchers on board and ensure we had all the expertise we needed, we added four new partners to the consortium (you can find the full list here). This made me appreciate our project officer, Adelina Nicolaie, even more, especially for her infinite patience and flexibility throughout the numerous changes.
Even with these unexpected challenges, we are still on track with respect to the original project plan. We’ve achieved breakthroughs and (cool) new concepts. The key technologies we are developing are swarms of resource-constrained sensors, and knowledge-based co-evolution. We rely on these to solve most of our problems: swarms can improve performance almost without limits by increasing the population size, and co-evolution can optimize design trade-offs that we cannot solve ourselves.
We are now close to achieving a feasibility proof for the system we are developing. Our first verification set-up (see below) bears an uncanny resemblance to a children’s swimming pool – because it is one! – combined with some transparent pipes, tanks and pumps so we can run the first very simple and tightly controlled experiments.
Even though we do not yet have a formal feasibility proof, we find that there is a lot of interest from industry in Phoenix technology. We have discussed possible applications with several companies and even intend to launch a spin-off company to further develop Phoenix-based products and services. Phoenix continues to be a very exciting adventure!
I would like to acknowledge all the researchers and students involved in the project, as well as Léon Verhoeven of our partner Antea, and of course our business development officer Jan Haagh who carefully organizes all possible and impossible details of this project.