A grocer becomes a robotics vendor
Colruyt Group, Belgium's largest grocery retailer, has agreed to sell 70 percent of the research and development activities behind its self-driving pallet trucks to KION Group, the German intralogistics company that owns forklift brands STILL and Linde. Colruyt keeps a 30 percent minority position through its Smart Innovation vehicle, and the associate will be carried using the equity method. The transaction is expected to deliver a positive financial impact of about 20 million euros to Colruyt in the 2026/27 financial year.
The structure matters more than the headline number. Colruyt is not licensing a feature or signing a supply contract. It is carving out an operating unit it built for its own distribution centers and letting a specialist manufacturer take majority control, scale it, and sell it across Europe, the Middle East and Africa. We read this as a retailer treating hard-won automation know-how as a commercial asset with an addressable market beyond its own four walls.
The software was the retailer's, the hardware was the vendor's
The autonomous vehicles trace back to a multi-year collaboration between Colruyt Group's Smart Innovation team and KION brand STILL. Colruyt engineers wrote the software intelligence that lets the trucks navigate crowded aisles and locate pallets with precision. STILL supplied the hardware integration and manufacturing discipline. The first autonomous STILL pallet trucks went live in 2023 and have operated inside two Colruyt distribution centers for roughly two years.
That split explains why the deal makes sense for both sides. Jo Janssens, Director of Colruyt Group Technics, said the company's engineers are developing robotics software that fuses AI and computer vision for future logistics operations. Koen De Vos, Director Supply Chain at Colruyt Lowest Prices, described the practical payoff on the floor, noting the self-driving vehicles navigate safely around people and locate pallets reliably. KION gains a codebase proven in live grocery logistics, and Colruyt gains a manufacturing and distribution partner it could never become on its own.
Why proven beats prototype in warehouse automation
Warehouse robotics is littered with pilots that never survived contact with real throughput. Systems that demo cleanly in a controlled bay stall when they meet mixed pallet loads, human traffic, and the relentless tempo of a grocery distribution center. Two years of continuous production use inside Colruyt's network is the asset KION is really buying. It de-risks the technology for every customer that follows.
For KION, this is a shortcut past the hardest part of the roadmap. The company already builds driverless forklifts, and it can now graft on navigation and perception software that has earned its keep in a demanding environment. The new KION Automation Center in Antwerp, supported by the Flemish innovation agency VLAIO, gives the combined effort a base to industrialize the platform for the wider EMEA market. Colruyt supplies the reference deployment that turns a sales pitch into a case study.
The build versus buy calculus, inverted
Most retail technology leaders frame automation as a build versus buy decision. Colruyt shows a third path. It built, it operated, and then it sold majority control while keeping enough equity to benefit from the upside and enough proximity to keep learning. The 20 million euro impact is real money, yet the more interesting return is strategic. Colruyt offloads the capital intensity and manufacturing burden of scaling hardware while staying close to the technology that runs its own supply chain.
This is the calculus we would push enterprise buyers to run on their own automation programs. If you have poured engineering budget into perception, routing, or fleet orchestration, ask whether that work is a defensible product or a private subsidy. The retailers who invented breakthrough logistics software have too often watched a vendor commercialize the same capability and then sell it back to the whole sector, including their direct competitors, at a premium.
What retailers give up when they let go
There is a cost to handing a specialist the keys. Once KION controls the roadmap, Colruyt's specific priorities compete with the needs of every other customer KION wants to win. A retailer that surrenders majority control can find its edge diluted as the same automation shows up in rival distribution centers. The 30 percent stake and the shared R&D center are Colruyt's hedge against exactly that erosion, keeping it inside the tent rather than reduced to another line on a price list.
The counterweight is focus. Colruyt is a grocer, not a robotics manufacturer, and sustaining a hardware business at scale demands supply chains, service networks, and capital that sit far outside a retailer's core. By partnering rather than going it alone, Colruyt frees its Smart Innovation team to chase the next problem instead of maintaining a growing fleet and a manufacturing operation. The discipline here is knowing which capabilities deserve permanent ownership and which are better monetized and released.
The signal for enterprise supply chain teams
The broader lesson lands on any CIO whose distribution network has quietly become a software shop. Automation IP developed in-house is now a tradable asset with real buyers, and the market for proven, deployed systems is far more liquid than the market for slideware. If your team has solved a hard logistics problem, there may be a KION-shaped counterparty willing to pay for the head start and share the future value.
We expect more of these carve-outs as retailers confront the true cost of maintaining bespoke automation. The winners will be the operators who treat their engineering output with the seriousness of a product organization, complete with defensible boundaries, licensing terms, and equity positions. Colruyt has drawn a clean line between the technology it wants to own outright and the technology it is happy to see scaled by someone else, and that clarity is the part worth copying.
