Sensible 4 is pleased to welcome Matias Koski as an advisor to the company. Matias brings a rare combination of experience spanning global technology companies, deep-tech scaleups, vision-based positioning, and industrial customer development. Before becoming CEO of Immersal, he built his career at Nokia and Microsoft, working across Europe and Asia before fully transitioning into startup leadership.
His decision to join Sensible 4 comes at an important moment in the company’s journey.
As autonomy matures, the challenge is no longer only technical. More and more, it is operational: how to deploy autonomous systems in the real world, keep them running reliably, lower the bill of materials, reduce maintenance burden, and create a solution that customers can actually scale across sites and fleets. In Matias’s view, that is exactly where the opportunity lies – and why Sensible 4’s new direction matters.
The winning approach is ‘plug-and-play’
For Sensible 4, this reinforces the company’s evolving focus: moving from engineering-heavy, custom deployment toward a more scalable, product-first model for restricted-area autonomy. Matias sees that move – especially the shift from public-road to off-highway and industrial use cases – as the right strategic choice. He believes the market is opening now, customers want solutions that work with existing fleets, and the winning approach will be plug-and-play wherever possible, with fewer engineers needed on site and a much clearer path to return on investment.
What attracted him most was the problem itself. Autonomous vehicles have already proven that many technical problems can be solved with enough money and enough engineering effort. The real bottleneck now is making autonomy practical and affordable enough for industrial operators. That means reducing unnecessary system cost, simplifying deployments, and focusing on customer outcomes instead of technical sophistication for its own sake. In Matias’s words, customers do not care how fancy the technology is – they care that it works, that it is reliable, and that the economics make sense.
Understanding industrial customers is vital
That perspective fits closely with Sensible 4. The company is building for operators in real environments: sites where uptime matters, weather matters, labor availability matters, and customers do not want to rip out existing fleets just to adopt autonomy. They want affordable upgrades, reduced vendor lock-in, and systems that can adapt to the realities of industrial work. Matias believes those needs will define the next stage of the market.
As an advisor, Matias will support Sensible 4 with experience in scaling a deep-tech business, understanding industrial customer requirements, knowledge of the mining sector, and insight into the role vision-based technologies can play in future autonomy stacks.
Below is our conversation with Matias on why he joined, what he believes the autonomy market is getting wrong, and where he sees the biggest opportunity for Sensible 4.
Q: Matias, for those who may not know your background, can you tell us a bit about your journey?
I started out studying material science and industrial engineering, and then followed a fairly classic global technology path through companies like Nokia and Microsoft. My work took me across several countries and roles, and over time I saw both the strengths and the limits of large corporate environments. Big companies can be stable and efficient, but startups offer something different: the excitement of building, the speed of growth, and the chance to shape the direction of a business much more directly.
That eventually brought me fully into deep tech. Having led Immersal, I’ve been deeply involved in building visual positioning technology – localizing devices through camera imagery. The field sits at the intersection of robotics, autonomy, and real-world deployment. What excites me most is bridging cutting-edge innovation with clear, scalable customer impact.
Q: What made Sensible 4 interesting enough for you to join as an advisor?
I like the problem the company is trying to solve.
There is still a lot of noise in autonomy, but if you look underneath that, the real question is not whether autonomous driving can be made to work in principle. In many cases, it already can. The question is whether it can be delivered in a way that customers can afford, trust, maintain, and scale.
That is what made Sensible 4 interesting to me. The company is moving toward a much more practical framing of autonomy: focused on operations, deployment, and productization. That is where the real opportunity is now.
Q: You’ve spoken quite directly about cost. Why is that such a critical issue in autonomy?
Because in industrial markets, the bill of materials is often the real bottleneck.
If your autonomous stack is too expensive, the return on investment disappears. And once the ROI disappears, it does not matter how impressive the technology is. Customers in mining, logistics, and similar environments need something that performs reliably, but they also need something that makes economic sense.
So if you can reduce the number of sensors, simplify the hardware, and lower maintenance cost, you immediately improve the business case. I think that is one of the most important areas where autonomy companies need to focus now.
Q: Does that mean you believe the future is vision-only?
I think vision will continue to become more important, absolutely. But in harsh industrial environments, I would not reduce the conversation to ideology.
Cameras are powerful. Vision-based positioning and perception are moving fast. But if you operate in fog, heavy rain, snow, dust, or other challenging conditions, you also need redundancy. Humans slow down when they cannot see. Industrial vehicles often cannot afford to slow down in the same way, especially in productivity-critical environments.
So my view is that camera-based systems are a major part of the future, but in demanding real-world operations you also need supporting modalities such as radar. The goal is not to build the most elegant theory. The goal is to keep the system productive and safe in harsh conditions.
Q: What do you think many autonomy companies still misunderstand about customers?
They talk too much about the sophistication of the technology and too little about customer outcomes.
Customers usually do not care how fancy the stack is. They care whether it works. They care whether it delivers clear value. They care whether they can keep it running, whether the maintenance burden is manageable, and whether the numbers make sense.
This is especially true in deep tech and industrial markets. You cannot sell complexity for its own sake. Reliability, uptime, and ROI matter much more than technical elegance.
Q: What are industrial operators actually asking for right now?
One major theme is that many customers do not want to replace their fleets entirely. They already have equipment in the field. They want ways to upgrade existing fleets rather than start over with completely new autonomous vehicles.
The second major theme is avoiding vendor lock-in. End customers do not want to be trapped inside a single OEM ecosystem if they can avoid it. That makes retrofit-friendly, flexible solutions very attractive.
And then there is labor. In many sectors, the work is difficult, remote, or unattractive to new workers. Over time, you will need more automated operations and new kinds of operator roles to manage them. That shift will happen gradually, but it is clearly coming.
Q: You’ve said Sensible 4’s move from public-road autonomy to off-highway is the right one. Why?
Because it is a more realistic and more commercially relevant market for a company like this.
Public-road autonomy is dominated by players with enormous resources. Competing head-on there is extremely difficult. But in restricted areas and industrial environments, the market is already opening. The use cases are clearer, the ROI can be more visible, and the deployment environments are more bounded.
That makes it a much better place to build a product-first company. If you can create something close to plug-and-play – not perfectly plug-and-play, because reality is never that simple, but close – then you have a much better chance to scale.
Q: How simple does the product ultimately need to become?
Simple enough that you do not need a large team of highly specialized engineers on site for every deployment.
There will always be some expertise required. That is just the nature of automation. But compared to the old model – where deep-tech companies often needed very heavy engineering support to deploy – the future has to be much lighter.
I do not think companies like Sensible 4 need to operate the way they did years ago. The same kind of work can now be done with much smaller, sharper teams if the product and deployment model are designed correctly.
Q: What do you personally hope to contribute as an advisor?
I know what it looks like to scale a business in practice. There are many things you only learn by doing, by making mistakes, and by seeing how customers behave in the real world. I know the mining industry and some of the relevant customer dynamics there. I bring experience in vision-based technology, which I think will play an increasingly important role in future autonomy stacks.
And finally, I hope I can help communicate why Sensible 4 deserves to be understood on its own terms.
Q: What do you mean by ‘its own terms’?
I think it is important that people understand that this is not a relaunch of the old Sensible 4 story. What I see now is a company that is becoming much more business-driven and product-driven. The market timing is different, the framing is different, and the focus is different. The technology has a proven base, but the company now has to communicate clearly that it is building for the market as it exists today – not for an earlier era of autonomy.
That message has to be repeated consistently. People need to understand not just what the company is leaving behind, but what it is becoming.
Q: Where do you see the biggest opportunity over the next few years?
In focused execution.
The technology will keep improving step by step, but the companies that win will be the ones that reduce cost, deploy early, learn from real-world operations, and build products customers can adopt without massive friction. In the near term, I think staying focused matters a lot – especially for a small team. Mining is an important starting point. Over time, there may be opportunities to expand into other industries or partner more broadly, but focus comes first.
Q: Final question: why now? Why is this the right time?
Because the market is becoming real.
A lot of the foundational technical work has already been done across the industry. What matters now is product-market fit, operational readiness, and commercial discipline. That is why I think this is a very interesting moment for Sensible 4. The timing is better for a company that understands operators, solves for practical deployment, and keeps the business case front and center.