You don't feel the gearbox until you push back. In a cobot's arm, the moment of resistance,the subtle, mechanical refusal before it yields,is the ghost in the machine. It's the line between a tool and a partner. AILOS Robotics is betting its entire company on erasing that line.
Founded in 2025 as a spin-off from Vrije Universiteit Brussel’s BruBotics research group, the Belgian startup is not building robots. It is building the muscles inside them. Its product is the R2poweR, a patented high-torque gearbox designed to be fundamentally backdrivable. This means a human can physically move a robot's joint with minimal resistance, a critical feature for safety in collaborative workspaces, wearable exoskeletons, or the delicate grip of a prosthetic hand. The company's seed round, a €3.5 million ($3.78 million) raise led by QBIC and High-Tech Gründerfonds (HTGF), is a wager that this specific piece of hardware will define the next generation of robots that work alongside us [HTGF, Feb 2026].
The physics wedge in a crowded gear market
The actuation market for robots is dominated by established giants like Harmonic Drive and Nabtesco, whose high-performance gearboxes are essential for industrial automation. Their strength is also their limitation for human-centric applications: they are stiff and non-backdrivable, optimized for precision under load, not for safe interaction. On the other end are quasi-direct drive systems, prized for their smooth, responsive feel but often bulky and lacking in pure torque density. AILOS positions R2poweR as a synthesis, claiming it combines the backdrivability and smoothness of direct drive with the torque density of advanced gearing in a compact, manufacturable package [TechFundingNews, Feb 2026].
The technical claim is that their Wolfram-based gearbox offers "extreme torque density for loaded joints" while maintaining "low backdrive torque for human-friendly motion" [TechFundingNews, Feb 2026]. In practice, this translates to robots that are lighter (needing less material and energy to move themselves), quieter, and inherently safer. It's a bet on a shift in robot design priorities, from isolated cages to integrated teams.
From university lab to factory floor
The path from academic prototype to industrial component is notoriously steep. AILOS is navigating it with the deliberate pace of a deep-tech hardware company. It has completed a minimum viable prototype with support from regional innovation agencies VLAIO and Innoviris and is now focused on pilot projects with robot manufacturers [MandA, retrieved 2026]. The target for the start of production is early 2028, a timeline that suggests the next two years will be spent on validation, manufacturing scaling, and locking in those first design partners [TechFundingNews, Feb 2026].
The founding team, Pablo López García and Stein Crispel, emerge from the BruBotics research environment, giving the company deep roots in the problem space but an unproven commercial track record. Their early validation comes from institutional investors with a history in European deep-tech and a notable accolade: being named a finalist for the euRobotics Technology Transfer Award 2026 [VUB TechTransfer, retrieved 2026]. The company is now hiring for roles like a Senior Test Engineer for gearbox validation, signaling the move from R&D to rigorous industrialization [AILOS Robotics website, retrieved 2026].
| Role | Name | Background |
|---|---|---|
| CEO & Co-Founder | Pablo López García | Spin-off lead from Vrije Universiteit Brussel’s BruBotics group [HTGF, Feb 2026]. |
| CTO & Co-Founder | Stein Crispel | Co-founder from the BruBotics research group [Crunchbase, retrieved 2026]. |
The integration gamble
The most significant risk for AILOS is not technical novelty, but commercial adoption. The robotics supply chain is conservative. Winning a slot in a humanoid or cobot design is a multi-year journey of testing, qualification, and relationship-building. The company will be measured against three immediate challenges:
- Design-win velocity. The planned pilot projects with manufacturers must convert into firm design commitments. Without public customer names, the traction narrative remains prospective.
- Manufacturing scale. The promise of "scalability for high-volume production" must be proven at cost points that compete with entrenched Asian suppliers [TechFundingNews, Feb 2026].
- Category timing. The company is betting on the simultaneous rise of multiple robot categories,humanoids, cobots, exoskeletons,each with its own adoption curve. A slowdown in any one could impact volume goals.
Their rebuttal is embedded in the investment thesis: that a European alternative for a critical, safety-focused component has strategic value, and that the underlying physics of their gearbox creates a tangible performance advantage that robot builders will pay for to differentiate their own products.
For now, AILOS operates in the anticipatory space between a brilliant lab result and a shipped component. Its question is not whether robots will need better gears, but whether the industry is ready to prioritize the quality of a handshake over the sheer force of a grip. The product is an answer to a cultural shift still in progress,one where the machine's first duty is not to overpower, but to understand when to yield.