Automation Unwrapped: Simon Delavalle, Chief Technology Officer, Veolia Nuclear Solutions UK

In this episode of Automation Unwrapped, Chris Eastham interviews Simon Delavalle, Chief Technology Officer at Veolia Nuclear Solutions to provide a glimpse into the remarkable projects and innovations revolutionising how we solve complex challenges by harnessing the power of automation. 

They discuss the use of robots and automation in the nuclear industry, particularly in the challenging task of cleaning up the Fukushima nuclear disaster site. Simon shares his experiences in designing robotic systems to navigate hazardous and unknown environments, and how artificial intelligence is transforming the industry. 

Chapters

• 0:00 Simon Delavalle, CTO Veolia Nuclear Solutions
• 10:17 M

Transcript

Chris Eastham: Hello, I'm Chris Eastham, a partner in the tech sector team at Fieldfisher and welcome to the Fieldfisher Automation Unwrapped podcast. In this series, my colleagues from across the tech sector team speak to their clients about the world of automation and what the future holds.

In this episode, I speak to Simon Delavalle, Chief Technology Officer at Veolia Nuclear Solutions, about his lifelong interest in solving problems with the use of machines, how his team and their robots are solving the dangerous challenges of cleaning up the Fukushima nuclear disaster, and how he sees robots further integrating into our lives in the future.

We hope you enjoy the podcast.

Chris Eastham: Hi Simon, and welcome on behalf of Fieldfisher, and on behalf of our listeners thanks very much for joining us today to talk about your perspectives in autonomous things, just to open the conversation up, it'd be great just to hear a little bit from you about what your role is and how you got to where you are today.

Simon Delavalle: Yeah, Chris. Yeah, thank you very much for inviting us to talk about it. So I'm Simon Delavalle or Simon Delavalle in French. I'm the Chief Technical Officer for Veolia Nuclear Solutions, which is the nuclear branch of the Veolia company. I'm based in the UK. I work with a team in the UK that is focused on special machinery and robotics, where I've been for more than 10 years now, nearly 14 years coming on. 

Chris Eastham: And why did you get into technology and, in particular the, you know, the, the nuclear industry? What excites you about your role?

Simon Delavalle: Yes. I think I've always had an interest in machinery and moving solutions, integration of a lot of different type of engineering um, so other carrier at Medtronics. And I work in robotics previously to build up even in the medical industry. So different industries on packaging industry, et cetera. Engineering technologies – that is having a real impact on various peoples on the way the world works.

Chris Eastham: Awesome. And we're seeing a big uptick in interest in AI. Certainly 2023 was the year of AI. And I think 2024 will probably be the year of finding interesting use cases for that technology. But when we're talking about robotics and autonomous things in particular what use cases you finding within the robotics space and in particular for intelligent robotics at Veolia?

Simon Delavalle: Yeah, so AI is an interesting one because it's actually been there for a long time. I remember studying it at university, but you're right that the birth has been created very strongly as it's a lot more accessible. As of last year, with a lot of the companies releasing AI technology to the general public.

In terms of impacts of the robotics and robotics technologies, there's definitely a push towards autonomy and autonomous systems now. So for a long time, what we call automation. So repeating actions have been programmed, has been used very effectively. But AI pushing the machineries and the robotics towards more autonomous decisions and on trajectory on tasks to be performed really opens a lot of new ways to think about how technology can adapt to project that we thought would be too difficult for technological solutions. So we're exploring applications that we not thought about before, like sorting some nuclear waste that was too difficult to be sorted by normal traditional machineries or having robotic platform been deployed in environment that was definitely too difficult before to be programmed in advance. For example, we spend many years in the last decades planning some projects such as the project to be deployed into the Fukushima damage reactor.

On that took a lot of careful planning and analysing data that was not always available in the best forms. On now we'll see these type of projects with AI and autonomous probably become a lot easier to plan and to adapt as the project goes along.

Chris Eastham: Yeah. I'm aware of your work in relation to the Fukushima nuclear cleanup. Are you able to tell us a little bit more about that and may, perhaps some of the challenges you've been facing or the learnings or developments that have come out of that in the robotic space? 

Simon Delavalle: Yeah, of course.

Yeah. It's a very exciting and challenging project which is what we like. The purpose of the project that we are working on for Fukushima is to be able to deploy sensors inside the core of the disaster. So as many people will know there's been a meltdown of the Raptor after the tsunami in 2011 on the states of the fuel, which is the most productive elements of the reactor. It's not well known on no decoctioning can be planned until this is fully characterised. 

So the project we've done with a client in Japan, which is Mitsubishi Heavy Industry is to conceive of a new machine that can be deployed in that extremely challenging environments to have various sensors to characterise where the fuel debris are, what is the nature of the radioactivity over there.

And there are many different challenges, but I think in terms of robotics, the two biggest challenges are the unknown environments, as you can imagine, after an explosion, no drawings or no previous mapping will tell us what it is really when we go inside it. On the second biggest challenges is the access to that area that has to be done through a very small hole relatively to the size of the machine. So it's a tunnel across the shield that is about four meters long on half a meter diameter. So we have to do a lot of operations through small apertures and on obstacles in that environment.

And for robotics, that's quite challenging because we can plan only to a certain point in advance. And this is why we have a machine that's a mixture of automation and collaboration of a human operator to plan some of the next steps.

Chris Eastham: So that's really interesting on how robotics is being used in specifically in, in that context. How do you see automation changing our lives more generally outside of the nuclear context?

Simon Delavalle: Yeah, I think so. I think in terms of automation, our lives have been changed already quite extensively.

What's very interesting is the advance, as you've mentioned before, of AI, on to a certain effect of ways to replicate autonomy of technologies. I think in terms of autonomous technologies emerging now, it definitely will give us a more natural and easier interaction with technologies in everyday life.

So whether it's at work or at your life outside work. On this will enable people to do more on different things. So in terms of work and engineering on how the environment we work in, I think the type of design analysis prediction we'll be able to do will be done in different ways thanks to artificial intelligence.

I also think on a personal level, it will enable people to focus on skills that make human people more valuable to remove a lot of the nitty gritty as automation has done already, but it will be going even further now with autonomous systems.

Chris Eastham: And then within the nuclear industry itself. How do you see automation changing the way that we currently operate?

What do you see for the future of the industry?

Simon Delavalle: Yeah, I think there's more and more now confidence in these technologies to be deployed autonomously on nuclear sites where the question was always, how can you make sure that the tool you've programmed is going to adapt to environments you're going to see and what we've started seeing for the last two years, and I'm sure what we're going to see more in the future years is that there'll be more robotic platform that can access areas that are too dangerous for people at a much lower cost. 

So it gives more accessibility to technological solutions that would have taken too much time to either develop a program or plan, and was a search prohibitive. I also think through these mediums, we'll have a lot more access to digital data – digital data is the backbone engineering's on technological solution, and whether it is for autonomous solutions to be able to be guided in that environments or for planning and creating the right solution with sufficient data to guarantee the success of accessing some of the most hazardous area on the nuclear sites.

Chris Eastham: In the nearer term, if I had to ask you for one prediction for the next sort of 12, 18 months, what would that be?

Simon Delavalle: I think we're going to see a lot of the buzz that's surrounding AI that I find quite confusing, I think, for the general public, filter out into real application for industrial point your soul solutions.

So I think we all heard about ChatGPT on the natural language model that's been used. I am pretty sure this is going to be used to communicate the interface a lot more efficiently with technologies. So programming robots in a more natural manner or. Working with software in a more natural manner.

So a lot of people that don't have the skill set to use technologies that they cannot use today. And that's going to be very powerful. I also think we're going to have a lot more what has been seen as sci-fi robots being deployed on more industrial applications, quadruped robots working like dogs around it.

There's a new immersion trend on humanoid robots. So it's going to make the industry think differently about what tools can be deployed that was not conceivable before. 

Chris Eastham: That sounds pretty exciting. I'd certainly take that bet on seeing more, use cases and easier interfaces, machines.

The quadrupeds I'll be keeping an eye out for. Simon, that's fantastic. It's been really interesting to hear from you today. So thank you very much for joining us. 

Simon Delavalle: Thank you very much. It's been very exciting to be able to talk about it. 

Chris Eastham: Thank you for joining us for this episode of Fieldfisher's Automation Unwrapped podcast. If you have any questions about today's update, don't hesitate to reach out to us. If you found it useful, do make sure to give us a like or review on your podcatcher of choice. Thanks for taking the time to listen, and we'll see you next time.