Introduction and Background

Antonio Marsico:
Welcome to the next episode of Zacco’s series of Fireside Chats. Today we are discussing the role of intellectual property in a rapidly evolving technological landscape. Autonomous driving technology. The car industry has seen incredible innovations in recent years and elements of autonomous driving in decision making are already implemented in many of the vehicles currently on our roads.

Autonomous systems like self-driving trucks or specialised farming equipment are already successfully deployed, and deployed in controlled environments like mines, ports or warehouses. These environments have predictable Pre-mapped routes, they have no public traffic, this makes the challenge significantly simpler than public roads.

Scalability and adaptability to diverse real-world scenarios remain key challenges for personal vehicles, and the timeline for widespread general purpose personal vehicle use also depends on governments and consumers who must be confident in their safety and the legal framework.

To discuss all these aspects, I’m here today with Joakim Frid IP Strategy and Portfolio Manager at Zenseact. The company was born as a tech joint venture in 2017, it took its name in 2020 with a clear mission to make transportation safer.

And your guiding principle is: Every mile should be safer.

You’re leading the charge of autonomous driving, but your strategy is brilliant. You’re not waiting for the future. You’re taking the most advanced software, AI, algorithms and sensor technology, designed for full self-driving, and deploying them in today’s advanced driver assistance systems, in our current cars by the way.

So, with the technology moving so quickly in this field, the primary focus of our discussion today is how research and development can be protected by intellectual property when digital technologies are so advanced, the developments are faster than the IP protection timeline.

Welcome Joakim and thank you for joining us today. Can we start by finding a little bit about you, your background as well as the Zenseact mission?

Joakim Frid:
Thank you, Antonio, yeah. My journey has been kind of a mix between engineering and intellectual property. I started to work as an engineer and after a couple of years. I entered the IP industry in 2002, as a patent attorney, and I stayed there for a couple of years before I jumped into the, Startup Ipendo, a company within, Intellectual Property services. And I must say that was a real experience to build a company like that, and all of the innovation around it. I stayed there for almost ten years before I switched over to the industry. Then I joined, Volvo Global Trucks Technology in the role as Global Head of IP Strategy and Portfolio Management.

Since 2018, I have been working in my current role as IP Strategy and Portfolio Manager at Zenuity, where I started, which became Zenseact in 2020. Zenseact is a fully owned subsidiary of Volvo Cars and even when we have been growing very very fast, we have tried to keep the startup spirit within the company.

And our mission is simple, but ambitious, we want to make traffic safer and ultimately eliminate accidents through advanced driver assistance and autonomous technology. To do this, we are working in the cutting edge of Deep Tech. and we are developing software and systems that enable autonomous vehicles. Many of our solutions are already out there in Volvo cars and Polestar cars and within the short term, we are focused on reducing near accidents and collisions. But of course, the long-term target is to enable the safe adoption of autonomous driving.

“Making every mile safer” through autonomous driving

 Antonio:
So, Joakim you made it very clear that Zenseact’s goal is to eliminate traffic accidents, through advanced driver assistance but also to get to full self-driving capabilities. Now what type of challenges do you still need to overcome in encouraging adoption of completely autonomous cars on public roads?

Joakim:
I think, autonomy is already well accepted in many places. You mentioned it already with the construction sites, agriculture, mining, and these controlled environments, and if you look around us, we also have a lot of robotaxis up there already, especially in the US and China, and some other regions around the world. We also have these highly advanced driving systems in our cars already, and they are evolving rapidly now as well. But I think the biggest challenge we have in front of us is scalability of autonomous technology, because it must be robust and adaptable to a huge amount of different environments and complicated situations. And this is especially true, I think, for personal cars, that operate in a completely unpredictable environment many times.

Our technology is already in production. We can see it in the Volvo EX90 and the Polestar 3 already. And, soon to come, we have the new beautiful Volvo ES90 and Polestar seven, and also Volvo EX60 a bit further on. But expanding our use cases now also means that we must keep pace when it comes to developments and here the data is really key, because if we can collect large volumes of data, we can also train these AI systems, and machine learning algorithms to keep pace with the development and the scaling up that that is needed now, and to be efficient with all this handling of this data, we also need some capabilities in place, for instance data curation. We must be able to annotate our data, and here we use what we call pseudo annotations, which is mathematically labelling the data. This has normally been done by humans, but it is too slow and too costly, so instead we automate it. We also need to have self-supervised learning in place, which means that the algorithm will teach themselves and learn when they operate, which is really crucial for our own foundation models to develop. of course, managing all this amount of data also requires, retention policies and deletion policies, because we cannot keep everything.

Another interesting area is also synthetic synthetic data, and here we can see that we can many times start from the real data, and then we can make clones and variants of it as a synthetic or semi-synthetic data, which can give us even more scenarios, and especially the ‘corner cases’ that can be very difficult and rare out there in the real world. There is, of course, also an operational side of all this, and that is ‘Over The Air’ updates and, cyber security, because these systems need to be constantly updated, to evolve, and they also need to be protected from different threats from outside. And then, of course, it’s not only about technology, we must also reach the public trust, and, then even also the legalisation side, which must evolve in with the same pace. I think the best way here is to show Success, and prove that it works.

Antonio:
To make it safer, to make it comfortable for even the users.

Joakim:
Yeah, exactly.

Different forms of research & development

 Antonio:
So thanks Joakim, you spoke about enhancing real world scenarios and therefore you need a lot of data, and scalability is another challenge. Definitely public trust, and legislative frameworks are needed in order to get to where we want, or where you want. So what type of challenges now, I would like to ask you, you have in order to decide on what to protect and how to protect the development that you are doing at Zenseact.

Joakim:
Yeah. Almost all our R&D is homegrown, we only have a few licenses, in the background, and the pace of innovation in our industry is very fast and it’s sometimes it’s outdated already When a patent is granted, to be honest. Especially in the machine learning area. it is moving extremely quickly. I mean traditional timelines in the IP industry for it, take for example, the 18 months for publication is not very much of a help for us to get insights into what others do. Because it’s already too late by then. So we are more and more looking into research. This is more of an actual and kind of immediate source to understand what’s happening around us. This speed also makes this very challenging for us to maintain a solid long term IP strategy, because you never know what’s around the corner. And if you also add the geopolitical situation in the world right now, the uncertainties become even worse, I think.

Historically, IP has been seen very much as a backbone to take things and block competitors, and maybe generate royalties as well. I think today that’s changed rapidly, especially in industries like ours, where everything goes so fast, and, everything is developed with high velocity. Instead it’s more and more, IP becomes more and more a tool to establish and enable collaboration, rather than competition or confrontation with others.

As you know, at Zenseact, we believe that collaboration is key, because we cannot do everything ourselves, it will be too complex and too slow, and by working together, the whole industry can grow much faster. But collaboration also requires a balance, because all these companies need to align around something, but at the same time safeguarding their own competition and uniqueness. That’s where IP plays a vital role for us, helping us stay open where it makes sense, while still protecting, what matters most for us here.

We have developed a strategy for this together with you at Zacco, where we can guide our organisation to understand which assets we should share, and under what conditions, and which assets we shall protect or keep internally. This framework also makes us, as an organization, to be agile and secure in a very dynamic environment.

Then another interesting shift in our industry, I think, is the competitive landscape. It’s not as much about the automotive makers any longer, and OEMs, it’s more like chipset manufacturers and tech companies today, and they also historically have had a quite different approach to IP. I would say. So that’s also a new challenge for us and makes the game even more complex many times.

Today very much of our development focuses on the backend system. What we’re trying to do now is to use more and more advanced technology in deep learning and machine learning, to enable existing functionality. So it’s not that much about new features and functionality that you and me Will experience in the vehicles, It’s more how we enable the existing technology, in a completely different way. So that also means that our IP strategy tries to balance this mix of offensive, defensive and, collaborative measures.

It’s very much about protecting our investments while being open to partnerships and then drive innovations in that way.

Personally I see IP in fast moving tech, as an enabler, more than a barrier. It shall create opportunities for Collaboration and innovation and not slow them down. The real challenge, and the real art, is finding the balance between openness and protection so that, progress can benefit everyone working together.

Innovation as one of Zenseact’s core pillars

Antonio:
So, thanks Joakim for mentioning Zacco as a partner in defining your IP strategy, and it’s interesting to hear about your competitive landscape. So now you have the same ambitions of some tech companies. You are competing with some of them while you are collaborating with, with the same company. So this rapid pace of technology development in AI, in sensors, in software, is driving a complex chang in in the IP strategy from mostly confrontation, to collaboration. This is what I hear from you. You need to collaborate in order to enable the innovation to come.

Joakim:
Now talking about innovation, with the software advancing in this way, with the fact that your innovation will be anyway, forming a super critical set of applications, because our safety is in your hands, how do you maintain your innovation momentum? And also how do you mimic the imperfections of human beings when they drive the cars?

To start with we are fortunate to have an incredible skilled team, or teams, and we also have, 12 PhD students working with the research in the company. We have a strong collaboration with different universities, in Sweden, like Chalmers, KTH, LTH, but also abroad, for instance, MIT and Stanford. And on top of this, every year we bring in about 40 Masters Thesis students, so it’s also a fresh flow of ideas coming into the company that way. Our culture in the company is really encouraging experimentation, and even when it means failing sometimes, because we also learn from failing. And innovation is one of our three core values in the company. So, it’s not only a buzzword, it’s a priority, a clear priority in the company.

As a subsidiary of Volvo Cars, we can leverage a lot on their heritage in safety, but at the same time, be an independent tech hub of Volvo Cars. And that independency is also supported by an open culture and noncompetitive environment in the company, their innovation and collaboration. And to give you just an idea how strong this culture is, we had last year one in five employees contributing to our innovation pipeline, which I think is really unique for a company of our size. And I have never seen it before, in any company, to be honest. Another thing we are working on this is to shorten the lead time between research and product implementation. Because in our field there happens a lot in the research, and that’s something we can utilise and build on. But then you take the research results and transform it into a product, you must solve a lot of different challenges. Because our world is not that perfect, we don’t have unlimited amount of compute, the data is not perfect, and a lot of other hard hurdles that come up. Here we also try to have a balanced IP strategy. Of course, everything is not protectable here, but some of these small, important solutions to industrialise the research, can also be protected. Yeah. And you also mentioned human-like driving, and this involves very much adaptation. First we want to make driving smoother and more natural, but it’s also to complement the logical approach taken by a computer. Because that’s simply not how people drive in reality. And here we try to mimic very much human driving, by collecting data from human driving and training our algorithm. In that way we will solve different problems, but it also makes the system, the whole system, more intuitive.

Developing a flexible IP strategy

Antonio:
Let’s go back now to discuss how you develop your IP strategy, what type of ideas you decide to protect, and how you think you can capitalise on their potential.

Joakim:
Yeah. That’s a great question because building the right IP strategy is really about thinking beyond today’s deliveries and what will come in the future. I wouldn’t say the classic exclusive IP approach is dead But in fast moving tech companies, we also need a more nuanced playbook. For us, that means aligning the IP more closely with our business model and long term goals, to make sure that Zenseact can stay relevant in the industry over time. To achieve thus, we use what we call our three pillar model.

The first part here is alignment with business goals and corporate culture, and here our IP strategy must support Zenseact’s objectives, but also reflect our collaborative culture. And, I mean, we value openness and partnerships over confrontation, that also shapes how we manage innovations and IP and even open-source contribution.

The second part here is balanced protection methods, and here we try to apply the most suitable protection method for each asset, whether it’s patents, trademarks, trade secrets or it can be also be academic publication, Or open innovation. The goal here is to categorise our different assets and apply the right legal and technical and organisational framework to it. The last part is mapping assets to value creation. This means linking the different assets to the best possible value creation, and here we try to balance the offensive strategies with the defensive strategies. The offensive ones to make us competitive, of course, and the defensive ones to mitigate risks, very much. Here we coach our teams to recognise and protect high value assets for our long term success. For areas where we think patents make sense, we actively encourage the organisation with innovation programmes. We help our teams to identify innovations and protect them when possible and suitable, and here we also have a structured evaluation method, and criteria behind it.

I wouldn’t say the challenge is to find ideas, it’s more to build the right mix of portfolio, so we can support both the short term needs and the long-term needs. So in short, I would say that our IP strategy is very much about balance and I think it maybe also makes us a little bit a more interesting proposition than our competitors. Because if you lean too much toward protection, it will slow you down many times, but on the other hand, if you lean too much toward openness, you risk losing your edge.

Protecting innovation to share it

Antonio:
Very interesting, and there is a red thread here, which is the dilemma about external collaboration. So. you need to collaborate, you said it loud and clear, you can’t solve all the issues on your own, so how do you decide what can be shared and what has to be protected? So where do you drive, and how you drive, your effort in terms of enforcing your rights?

Joakim:
Yeah, that’s a great follow up question because I think that really sticks to the operational side of the dilemma. At Zenseact, as I said before, collaboration is essential for innovation. But we also know that we must be able to safeguard our core assets, and also for enforcing our rights, in the future, or in the long term competition. We often start by mapping our assets for collaboration versus Protection.

Sensitive solutions, then the key technologies are never disclosed in a shared environment, or an open environment. While more general technology and exploratory work, fits much better into this open area and in that way we can also gain from joint learning without risking what matters most for the company. Before we consider sharing anything, we always consider if we need a formal agreement in place or not. It can be NDA,or collaboration agreement, where weclarify ownership. It can also be the permitted use of the results, and, enforcement rights at the end. But we never want these agreements for the sake of it, because they tend to slow down things in long discussions and negotiations. So if we can collaborate, without agreements, we try to do so, but of course, it’s not always possible. We also train our employees, internally, in confidentiality and compliance, and we also provide tools to monitor our different environments, and we enforce boundaries when we’re moving between the open and closed areas. Many of the companies around us have huge portfolios. We need to make sure that we balance that with something in our portfolio as well, and protect things. But in short, I would say our approach is pretty simple. We are open where possible and we are closed where necessary.

Antonio:
You are trying to say that you make it simple but actually you are making simple something which is very complex.

Joakim:
Yeah, it is complex, but you’re right, we’re trying to make it simple, and we’re also trying to make it simple for our employees. Because if you don’t know, you can many times feel uncertain. ‘what I’m allowed to share and what I’m not allowed to share.’ When you when you go out and meet other people there.

Round up and close

Antonio:
So Joakim thank you very much for this fantastic look into Zenseact’s strategic DNA. Your technology will clearly play a role in the future of traffic safety and autonomous driving, I look forward to getting there. We have established that in autonomous driving, IP is no longer just a weapon for blocking, it’s an enabler for collaboration, and for innovation.

So your balanced three pillars, IP strategy aligning protection with business goals, mapping assets to value, and choosing the right measures is clearly key to accelerating development, while securing your competitive edge in this incredibly fast-moving landscape. Everything is fast, the pace is there, and you can’t block the pace, otherwise you will not do innovation. So, thank you for sharing Zenseact’s incredible journey, sharing your insight into how you are successfully managing the combination of development, protection, collaboration In, again, in a field where everything moves very fast. Zenseact, as a company is firmly at the forefront. Your focus on testing, development, prioritising passenger and pedestrian safety, your close collaboration with universities, with other partners, with governments, leaving me feeling confident that the future of self-driving vehicles is really in good hands.

Thanks Joakim, once again, and thanks everyone for looking at this episode of Zacco’s Fireside Chat, today on the role of intellectual property in a rapidly evolving technological landscape.