The challenge behind the systematic factor

Since May 2024, Kempower has led HETE, the Heavy Electric Traffic Ecosystem, a five-year programme funded through Business Finland's Veturi scheme for company-led ecosystems. Business Finland granted Kempower €10 million for the work and set aside a further €20 million for projects run by ecosystem partners. Tommi Rissanen, who heads the programme, also represents Kempower on RAIDE's steering group.

HETE was set up to remove the systemic obstacles to electrifying heavy transport. Charging power far beyond passenger-car levels. Grid capacity at depots, ports and terminals. A data layer that lets fleets, grid companies and charging operators work together. Trucks account for 27% of transport emissions in the EU, and by HETE's own account, no company can meet the energy needs of electric trucks on its own. Most of the work is early-stage research.

Two years of work have generated a lot of insight and at the same time a lot of questions to further investigate.

The cluster Kempower joined

RAIDE — the Radical AI Deployment Ecosystem — was launched in May 2026 by Demola Global, the Finnish innovation platform, together with Business Finland, which funds its early-stage operations. Its founding partners are Wärtsilä, Danfoss, Kempower, Bittium, Valmet, and the wellbeing services counties of Pirkanmaa and Ostrobothnia. Its stated ambition is for Finland to be a world leader in applied AI by 2030.

The mechanism is straightforward. Partners bring needs and opportunities they have identified; Demola works those into briefs and problem statements; the briefs go to universities as course projects and master's theses, and to startups as pre-pilots. Results return to the partners, who decide what to do with them.

"Juggling two innovation horizons at once is a daunting and resource-intensive challenge, and the openings AI is creating right now lie beyond the long-term horizon which is always the more difficult one" says Ville Kairamo, CEO of Demola Global. "RAIDE gives our partners a shared long term innovation horizon with resources and cross-industry co-creation with Finland’s industry leaders. That’s a superpower Finland has never fully used."

The exchange runs in both directions. Kempower joined for what the other partners and the research community could bring it, and arrived with a problem the rest of the cluster can work on.

"Charging heavy transport is exactly the kind of question that crosses industries — energy, logistics, data, public infrastructure," Kairamo says.

Creating value from two innovation horizons at the same time

"HETE gives us the challenges. RAIDE brings the companies, cross-collaboration and the leading edge of research to solve them," Rissanen says.

The two do not overlap, because they are made up of different people. HETE goes deep into one industry — truck manufacturers, depot and terminal operators, grid companies, financiers — who have worked on the same problems long enough to trust one another. RAIDE puts Kempower alongside companies and research groups it would not encounter in heavy transport. Kempower works in both, and takes something different from each.

Step back, and applying AI begins to resemble electrifying heavy transport. It crosses industry boundaries, it starts with the application rather than the technology, and no company can solve it on its own.

"That is why the same approach works for both," Rissanen says. "Give people a challenge worth solving and let them research it properly, and what they find turns out to be useful far beyond the case they started from."

A Crystal Ball that Reveals What Customers Think

Rissanen is direct about what Kempower expects from this, and it is not a feature list.

"We're not focused on getting product features or small improvements out of RAIDE collaboration," he says. "What we want, and what we're getting, is an early view of where AI in heavy electrification is going, before customers start asking for it."

HETE organises its own research into four fields: raising the power capacity of charging systems, finding new uses for data, supporting overnight, destination and on-the-move charging, and developing the charging technologies that will follow today's. The four areas Kempower brings to RAIDE cut across those fields rather than mirroring them.

1) Forecasting charging demand from data

Charging a fleet is a scheduling problem before it is a hardware challenge. Routes, shift patterns, weather and driver hours all shift demand across the day, and an operator who can see tomorrow's peak coming buys less grid capacity than one who cannot.

2) How fleets and the electricity grid interacT

Grid capacity at depots, ports and terminals is the hard limit on electrifying heavy transport. The open question is how much of that limit can be worked around by smarter behaviour on the charging side — spreading load across a site, responding to price and grid signals — rather than by new cable in the ground.

3) Measuring emissions and efficiency

Electrification is sold on a promise about emissions, and that promise has to be measurable per vehicle, per site and per kilowatt-hour before financiers, regulators and freight customers will rely on it.

4) Pushing decisions down to the device itself

A megawatt-class charger dividing power among several vehicles cannot wait for a round trip to the cloud to decide how. The more the equipment resolves locally, the better the system copes when the connection is slow or unavailable.

None of this replaces Kempower's own R&D. What it changes is when the thinking starts, and how often new ideas reach the people doing the work.

Research capacity without a laboratory

Through RAIDE, Kempower reaches research and cross-industry expertise and capacity it has little reason to build in-house, without committing the capital an AI laboratory of its own would require.

The clearest example is the thesis pipeline, and Rissanen is specific about where the topics come from.

"They come out of our own technology roadmap and the capability gaps we've identified there," he says. "Not from random matching based on who happens to sign up."

Kempower is considering startup collaboration on the same basis. There is also the effect on the company's own thinking.

"The same AI problems are being solved in different industries under different constraints," Rissanen says. "People who move between sectors see patterns you can't see from inside one industry."

That feeds into Kempower 2.0, the strategy the company published in May 2026, which states the aim of becoming one of the world's three largest providers of DC fast charging by 2030. On Rissanen's account, a position like that is difficult to reach without research partners outside the company.

Building a table for all

Kempower could have run an AI initiative of its own. What settled it was who else would be at the table: Wärtsilä, Danfoss, Valmet, Bittium and two wellbeing services counties.

Getting in at the start mattered just as much.

"We didn't join a ready-made structure. We get to define what happens in it," Rissanen says.

That is a different position from membership.

" The companies that came in at the start are not here to use something somebody else built. They are contributing, and all of us, from companies to Demola Global and Business Finland on a national level, are in this for the long run.", Kairamo says.

It is also why Kempower did not simply start something of its own.

"Applying AI is not an industry question. That is why the answer is a cross-industry cluster and not a separate project of our own," Rissanen says. "Finland being the leading country in applied AI by 2030 is a goal that only comes out of several industries sitting at the same table. We wanted to be building that table from the start."