Welcome to this morning’s session of the Science, Innovation and Technology Committee. I am very excited by our guests this morning for our innovation showcase. I will now hand over to Emily Darlington MP, whose showcase this is, to introduce the guests.

Thank you very much. Hello, everybody, and thank you for this opportunity to come and speak to you, and for me to bring our robot friend here to come and see you as well. I know that some members of the Committee are already familiar with Starship, but not everybody is, so I will give you a little bit of background about how we came into being, what we actually do and how we work in Milton Keynes and across the globe to deliver. As you can see on the video, we are the world leader in autonomous robot delivery from Mars to Milton Keynes. That is how we started out. People say, “Why are you called Starship?” Our founders are Ahti Heinla and Janus Friis. They were two of the co-founders of Skype. When they had finished with Skype, Ahti asked, “What shall I do next? I know, I’ll enter a NASA competition to invent a robot to go to Mars and collect space rocks,” as one naturally would do, of course. They didn’t win that competition, but they looked at the prototype that Ahti had built and thought that there had to be some other application for it. They looked at the logistics sector, an industry that still had not had its technological revolution, and they saw an application for what the Starship robot could do, so Starship was born. You can see on the video the little path of our robot starting to wend its way through lovely Milton Keynes. It rapidly turned into quite a few more robots wending their way through lovely Milton Keynes. We started out in Milton Keynes in 2018. We were founded in London by two guys with a whiteboard and an idea in a room with their little robot. It took a lot of local authority co-operation. We will come on to regulation later, but it has taken authorities to be bold and to push the boundaries a little bit to introduce technology into their communities. We started out in Milton Keynes with our very first commercial deliveries with the Co-op supermarket. We are now in 200 locations globally. We have driven over 9 million miles across numerous countries, completing more than 7 million deliveries. We make three crossings every second that we are in operation. That shows you the global scale of something that started out making a few deliveries in Milton Keynes, and where it has blossomed from there. We operate in a number of other UK locations, such as Greater Manchester, West Yorkshire, Northamptonshire, Cambridgeshire and Bedford. The reason we do that is that on demand is in demand. We all know that people have become accustomed to getting even small amounts of goods delivered to them quite quickly. That presents an ethical and sustainability issue about how we can do that as an economy and a society. It also creates a problem for business. When people want small amounts of things—relatively small baskets—the robot can take up to about three bags of shopping and deliver it to you. How do you do that and ensure that everybody in the supply chain benefits? That includes the actual suppliers themselves, the delivery riders and their wages and the people who actually make the goods. How do you ensure that everybody benefits? The last mile is notoriously the most carbon-intensive and expensive of the entire supply chain. Getting it right: as Emily said, the robots are loved. Part of that is the approach that we have taken to how we do this. The autonomy level really matters because how the robots behave around people is dictated by the level of autonomy. When our robots started out in Milton Keynes in their test phase, they were followed by people driving them with remote controls because, as with any sort of AI, the robot has to learn the world around it. It didn’t just come into being in year one and understand what a road crossing was, or a guide dog or a wheelchair. It has taken a long time to teach our robots what the world is. We are 10 years old. We only started making deliveries in 2018. It is a curve. You start off with a few deliveries, but as soon as you get the autonomy right you really get going. We are at level 4 autonomy, which means that we are 99% autonomous. We have remote assistance back-up in case the robot encounters a novel situation, but by and large the robots are operating autonomously most of the time. Because of that, the cost per delivery is lower, which is attractive to retailers because you are not having a lot of human intervention to help the robot do its job. As I have already touched on, the autonomy helps with social acceptance, too. The engagement and the accessibility work that goes alongside this is incredibly important. You cannot just turn up to an area and go, “Hey guys, you’re having robots now. Get used to it.” We forget in Milton Keynes sometimes that actually when you go to new areas people have not seen the robots before. They have gone from novelty to normal in Milton Keynes over the space of time that we have been there. People just accept them, and rely on them, as part of everyday life and everyday service, but we work really hard on that. It is about engaging with the council. We never operate anywhere that the council doesn’t want us. We engage with local accessibility groups. We have an accessibility advisory panel. All the work we did in Milton Keynes in the beginning forms part of our engineering and autonomous driving pipeline to teach the robots how to behave more sensitively. For example, the accessibility advisory panel has people who have low vision, are blind or are in wheelchairs, which has helped us to change the robot behaviour and develop our autonomy. It all goes hand in hand. Importantly, we have had to have buy-in and boldness from local authorities because we exist in a regulatory grey area. Our robots operate primarily on pavements. We cross roads but, as you can see, this little guy beside me wouldn’t fare too well on the M62 during rush hour. We have had to work really hard to socially integrate, and to get permission and buy-in from local authorities. To be honest, Milton Keynes took a bit of a risk when this first started out. No one had let these robots operate commercially before in their cities. Milton Keynes was not a city at the time. When Milton Keynes bid to be a city, the robot delivered the bid for them. We like to think that we played a little part in that history. I have touched on the benefits, so I will not dwell on that too much. We are all familiar with those. It drives sales because people like robot delivery. It also sings. It isn’t going to sing today, unfortunately, because the wi-fi is not quite good enough, but it can sing when it delivers to you, which is quite nice. There is increased productivity too, because it frees people to do more productive and profitable work in serving that growing market. We have had independent surveys around our societal and social acceptance. Councils said, “Okay, we’re going to let you do this, but we’re going to check that residents don’t hate it during the trial period.” The response has been overwhelming. Between 75% and 93% of people responding to a council survey—this was a council survey; no one is ever nice to the council, are they?—which included residents, not just customers, said that they liked the service, and they wanted it to stay. On top of that, a third of our customers say that they have a disability or live with someone who does. We are not just working hard to operate well around people with disabilities. We are actually serving a need in the community too. We firmly believe that tech is for everyone. We are across all of Milton Keynes now. We operate in all types of communities. This is not just tech for the privileged few. We want to make sure that everybody has access to it. Using things like our Starship Schools Programme, you would be amazed—or perhaps not—that it is much easier to get kids interested in STEM if you take a singing robot into school and let them drive it about a bit. It works really well. For the environment, our robots are zero emission. They have wireless charging stations, so they can plug themselves in and charge themselves up when they get tired. If you compare it with traditional combustion engines, we estimate that we would have removed around 350,000 kg of carbon from the atmosphere by replacing car trips. Our customers tell us that the vast majority of what we do replaces car journeys and takes cars off the roads. You can see our timeline on the video. That is our journey from starting out in Milton Keynes. People in Northampton saw it and said to Milton Keynes, “How do we get robots? Can we have some robots?” The good people of Milton Keynes then basically turned into ambassadors for Starship. We have people in Milton Keynes council who are probably more expert on robot behaviour than I am at this point. They went out and spoke to people in Cambourne, in Bedford and Leeds to try to help de-risk it and say to people, “It’s okay if you want to do this. People like it; give it a go.” That has helped us to really grow. As you can see, we didn’t launch much in 2024. There is a reason for that. The UK has fallen behind in this part of regulation. The Automated Vehicles Act was very progressive and very good, but it did not get our little robots. We are currently seeing in the devolved powers White Paper that micromobility is about to get its turn in the spotlight, but robots are not really included in that. If you speak to officials at the Department for Transport, as local authorities tend to do, “Hey guys, is it okay if we do robots?”, the response is, “Well, yes, maybe. It’s a bit of a grey area, but the robots could be classed as a carriage under the 1835 Highways Act.” With all the will in the world, I don’t think the good folk who were drafting the 1835 Highways Act necessarily had in mind a little robot that was going to operate and deliver people’s shopping autonomously in 2025. I could be wrong, but I don’t think that was necessarily the intention. That is why there are fewer robots at the moment than there might be. In Finland, they did a lot of trials—much like Milton Keynes did—but they took the trials from local authority level and immediately regulated nationally to create a national framework. We now partner with the largest retailer in Finland. We deliver as far north as Lapland, so the robots have had to learn a lot about snow. As you can imagine, just as a side point on the autonomy, if robots learn the world, and then all of a sudden the world changes because it is completely covered in snow, there is a whole new set of things that they have to learn. That was our learning last year in the Finnish winter. There is support for that sort of regulation. Again, I won’t dwell on this, but in areas where people are familiar with the robots, as you can imagine, people like the robots and there is support for Government regulation to ensure that more people have the benefit of this. Finally, where next? Progressive regulation. This comes at a very apt time, given that the AI opportunities action plan was published yesterday, which we really welcome. As much as focus is on generative AI and the potential economic benefits for the UK from generative, our call would be, “Please don’t forget embodied AI.” We are taking AI and putting it into hardware. We are going out into the world. In some ways that is much more difficult than something that exists online. Don’t forget about embodied AI. Cross-departmental action is needed and would be very welcome. It could be in the form of a national framework or devolved powers. We firmly believe that local authorities are well placed to understand what works in their communities and to enable that, much as Milton Keynes and other local authorities have done. Maybe it fits in the micromobility space with the regulation that looks to be passed around e-scooters. We are celebrating innovation here and we are so proud of the start that we made in 2018; it was a springboard that launched Starship and allowed us to go on to great things, but we want to stay here and do more. It is becoming more urgent as more countries develop this regulation and the UK does not. Thank you. (Applause.)

Thank you so much, Emily, for introducing Starship robot. Thank you, Lisa, for that fantastic presentation. I think I am right in saying that this may be the first time that this Select Committee has had a robot for a witness. For the benefit of Hansard, the robot was communicating with us with various lights. I am not sure exactly what they meant, but they were flashing. It is certainly a huge testament to the innovation that is taking place in Milton Keynes. The Committee has really heard your key messages on regulation and embodied AI. Thank you so much. Thank you also to the Clerks of the Committee for enabling the robot to be with us today.