 Machine learning technology can be combined with other data technologies and one area where this is being developed is in the realm of cyber security. At Cardiff University, a research team led by Professor Pete Bernab is combining machine learning and cyber security to develop detection and prediction models which can identify different categories of cyber attack and automatically block them. The focus of our research is the use of machine learning to keep people safe in a digital world. We're all using digital technology in our everyday lives, iPhones, iPads, computers, and they're all interconnected and humans can't possibly monitor these 24-7. It's just not possible. So if you imagine you had an AI or machine learning sidekick that was working with you and was able to advise you on when bad things were happening and what you should do about it that would really, really help us manage the situation. So take ransomware for example. This is something you can get onto your computer and encrypt your files very, very quickly meaning you don't have access to them. So machine learning can actually spot this in real time and act against it to block the ransomware and protect your files meaning less get encrypted. Now this happens 24-7. There's always attacks happening on digital networks so machine learning can work all around the clock, doesn't eat, doesn't sleep, never gets sick. So it's a really useful aid. The team's research has many applications. Cybersecurity is going to affect every person in everyday life. This is because more and more digital technology is used in transportation systems, in banking, in hospitals and healthcare. There's no getting away from it. So to think you're immune for a cyber attack, that's just not likely. So it's really, really important that everybody considers this. A major issue would be if a cyber attack happened on a hospital. So cybersecurity in hospitals, cars and future connected vehicles, these are all areas where we need to think about the cybersecurity of the technology before we roll it out. We're currently developing AI assistants, if you like, in the lab to work on detection and responding to cyber attacks and advising people on how they might like to respond. There's a massive commercial market for this and this is where we see it going in the next five years. The translation of research from the lab into practice, monitoring people's digital networks and advising them on how to respond in real time. In our research labs at Cardiff University, we're building representative models of real world systems. So from smart cities, connected vehicles, manufacturing systems, we're trying to make representative systems that ultimately we can come and learn on. We can launch an attack on that smart city. So lights will go off, transportation systems will start moving, driverless vehicles will actually start swerving across the road. We can see the impact of these real attacks in real time and then understand how to develop solutions to make them more secure. Another example is our connected vehicle. So this is a driving simulator game, computer game, but behind the scenes is a real world operating system that would be used in a connected vehicle. We can manipulate that, we can attack it and actually see the impact of it in the game in real time. And then again, we can learn how to secure it so that we can deploy connected vehicles of the future with security built in. What skills and qualifications do you need to get into the field of cybersecurity? Cybersecurity careers are changing all the time and that's in response to the changing digital environment that we're working. So if you go back 10 years ago, we all worked on laptops or computers within protected networks. Now we've moved to online social media, mobile phones and laptops that we take with us everywhere we go. So the cybersecurity careers have moved to respond to securing those types of environment. Now actually, what we're talking about next is connected roads and connected health and everywhere you go there's something that's monitoring you and trying to understand your behavior and how to better engage with society. And that again, changes the needs and the skills for cybersecurity. Particularly, it's going to mean a lot more interdisciplinary requirements. So people from sociology, psychology as well as the economics and the engineering and the computer science elements of cybersecurity is going to become much more of a focus on different skill sets contributing to new cybersecurity demand. We're really looking for people who just have an interest in the topic of how things work. If you think about it, cybersecurity is protecting a digital world. A digital world is made up of data and devices, computers, mobile phones, laptops and ultimately understanding how they work means you can understand how they can be attacked. If you understand how they can be attacked, you can understand how they can be secured. So really, there's no particular skill set. There's genuine interest in cybersecurity and technology. If we can have that, that would be brilliant for us going forward to future roles. There are different career pathways for you to progress through. So for example, you might follow a technical route and understand how to secure the technology side. I would say computer science is a core topic. If you can understand computer science, the code, the technology, how things work, this is a really important baseline. Maths is a nice addition as well. But it's also a route through that understands how people use technology. So people from sociology and psychology backgrounds, if you're really interested in how people work, there's a route for you to. There's lots of different ways into cybersecurity because it's such a practical and vocational subject. What really matters is you have a passion for it. Can you explain in more detail what a cyber attack on an autonomous vehicle would look like? I can, yes. So autonomous vehicles, as you can imagine, they're full of digital sensors. They get data. They receive data. They make decisions based on those. Now what that means is that because they connect to the internet, people can send in incorrect data, confuse the sensors. Then, we're often seeing, is attacks on the vehicles, robberies, etc. So that's a huge problem. Another thing that everyone's, I suppose, more versed in is GPS. So you can attack GPS systems, send a vehicle in a completely wrong direction. So a cyber attack could be just open the doors, unlock the vehicle, for example? Absolutely. I think a few people have said recently that crime has moved online, which has effectively similar crimes but now digitally enabled and effectively cybersecurity is just trying to prevent that. And then I suppose you talked about autonomous vehicles and all the sensors, but we all have those sensors in our houses. We have the temperature sensors, but the alarm systems, cameras. So do you think that that threat of cyber attack will put people off automating their homes? Well, as you say, we have a lot already in our homes. We've got Alexa's. We've got digital cameras. We've got phones and laptops, et cetera, which are all connected to the Internet. And I think the one thing that we need to get across in cybersecurity is to be aware of the threats that that poses. Yes, it's awesome. It's great. We can see our doorstep. We can see all around our house. We can get things all automated by digital assistance. Amazing stuff. However, we've got to think of it in terms of, you know, if you saw a dark alley, you wouldn't naturally just walk down that alley. You'd think, hang on a minute, where's the risk? Where's it coming from? What am I exposed to? And you'd act accordingly. And cybersecurity is really just making people aware of those threats in that digital environment. So it's the main problem you see, not that people are worried about cyber attack, but actually they don't think about it at all in the context of cars or their homes. I think that's exactly the problem. It's a classic technology problem that we're now seeing again with, for example, AI. Technology evolves. People adopt it and they use it because it helps them. They get value from it. But technology is always advancing quicker than cybersecurity. And it's always an afterthought. And so this is where we really need to get to that point now where we think, hang on a minute, what are the risks of using it as well as the opportunities and ensure that it's secured as we go? Well, thank you very much, Pete. You're welcome. Thanks, having us.