A Swedish industrial robot can now pick more than 2,000 items in an hour. That speed claim, from Lund-based Cognibotics, is the headline number for its HKM1800 robot arm [Cognibotics]. The less flashy, but potentially more valuable, product is the software that tells the arm precisely where it is in space. The company just secured a €6.5 million Series A to sell both [Cognibotics, 2026].
Cognibotics is an academic spin-out, founded in 2013 from Lund University’s RobotLab. Its bet is that advanced motion control, calibration, and novel robot design can solve two core industrial problems: accuracy and programming complexity. For a manufacturer welding a car chassis or a logistics center picking millions of parcels, a millimeter of error or hours of downtime for reprogramming costs real money.
The Wedge: Calibration as a Product
Cognibotics does not just sell robot arms. Its initial wedge is CogniCal, a calibration system that compensates for geometric deviations, elastic effects, and payload influence in industrial robots [RobotToday]. In simple terms, it makes a robot more precise by teaching it its own physical imperfections. This is a software and IP licensing play aimed at original equipment manufacturers (OEMs) and large end-users.
The commercial proof point is an order worth approximately SEK 16 million (about $1.5 million) for CogniCal from an unnamed global robot manufacturer [Mynewsdesk]. For a deep-tech startup, a seven-figure order from an industry leader is a traction signal that outweighs a dozen pilot projects. It suggests the technology is moving from lab validation to production line integration.
Building the Full Stack
Calibration is the entry point. The company's portfolio has expanded to include a proprietary programming language and a physical robot platform, aiming to control more of the automation stack.
- Juliet&Romeo®. This is a low-latency, real-time multitasking robot programming language and virtual machine [Cognibotics]. The claim is that it combines the reliability of traditional programmable logic controllers (PLCs) with the flexibility of a language like Python. For engineers, the promise is faster development and safer operation.
- HKM1800. This is the company's flagship robot arm, designed for high-speed picking in logistics and manufacturing. It uses lightweight materials and a low-inertia design, with 80% of the manipulator's weight placed in the stationary base for rapid acceleration [The Robot Report, 2025]. The stated pick rate of over 2,000 items per hour is a key marketing metric against rivals [Robotics 24/7].
The stack approach allows Cognibotics to pursue multiple revenue streams: licensing calibration IP to OEMs, selling its HKM1800 arms to system integrators and end-users, and licensing its software tools.
Funding a Decade of R&D
The company's funding history reflects its long gestation as a research-driven venture. Early support came from Lund University's own investment arm, LU Innovation, and Swedish innovation agency Vinnova [Cognibotics]. The recent capital injections are larger and aimed at commercialization.
| Date | Amount | Lead Investor | Purpose / Note |
|---|---|---|---|
| 2024 | $5.7M (approx.) | Newton ApS | Undisclosed round for market adaptation and product packaging [Fairpoint Capital] |
| 2026 | €6.5M | European Innovation Council Accelerator | Blended finance (€2.5M grant, €4M equity) for HKM1800 industrialization [Cognibotics, 2026] |
Newton ApS, a Danish investment firm, appears to be a major backer, leading a $5.7 million round in 2024. The 2026 Series A from the European Innovation Council (EIC) is notable. The EIC Accelerator is a highly competitive, equity-based grant program for deep-tech startups. Winning it provides not just capital but a stamp of technical validation from the European Union.
Traction Beyond the Lab
Customer names are still sparse in public reports, but the partnerships and projects cited point to a commercial footprint. Logistics firm Nowaste ordered a second robot-picking cell after a successful first deployment [Ideon Science Park, August 2025]. E-commerce company Boozt is collaborating on a pilot using the HKM1800 to improve AutoStore efficiency [Ideon Science Park, August 2025]. In a different vertical, cobot maker Standard Bots selected CogniCal for its next-generation AI robots [Ideon Science Park, August 2025].
The company also contributes to research consortia like the CAISA project, a Vinnova-funded effort developing a pediatric heart-surgery research demonstrator, which brought in SEK 10 million (about $1.1 million) in project funding [Robotics and Automation News, 2026].
The Founder's Forty-Year Bet
The technical depth traces back to founder and CTO Klas Nilsson. He spent time at industrial giant ABB before becoming a researcher and associate professor at Lund University, where he heads the Robotics and Semantic Systems group [IEEE Xplore]. Sources cite over 40 years of experience in the field [Alessandro Genua - Scuola Superiore Sant'Anna | LinkedIn]. This is not a first-time founder chasing a trend; it is an academic operator commercializing a lifetime of research. The company's public profile suggests a team of 11-50 employees [PERPLEXITY SONAR PRO BRIEF].
Where the Wheels Could Come Off
The risks here are classic for capital-intensive deep-tech. First, the sales motion. Landing a major OEM order is promising, but scaling a licensing business with large, slow-moving industrial customers is a different challenge from selling SaaS. The sales cycles are long and the procurement hurdles are high.
Second, competition. While no direct competitors are named in the sources, the market for robot calibration and motion software is not empty. Established automation suppliers like Siemens, Rockwell Automation, and the robot OEMs themselves have vast software portfolios. Cognibotics must prove its algorithms are not just better, but sufficiently better to justify the switching cost for a manufacturer with a factory full of legacy systems.
Finally, capital intensity. Developing hardware and complex real-time software burns cash. The disclosed funding total is approximately $9.93 million [Tracxn, 2025]. The €6.5 million EIC round will fund the industrialization push, but the path to profitability in hardware-heavy robotics often requires further, larger rounds.
Cognibotics has cleared the first major hurdles: technical validation from the EU and a flagship OEM order. The €6.5 million from the EIC Accelerator and Newton ApS is now on the table to prove the business model. The next twelve months will show if the company can convert its SEK 16 million calibration deal into a repeatable enterprise sales pipeline, and if the HKM1800's 2,000-pick-per-hour speed can win against established logistics automation giants. For a robot to be useful, it first needs to know exactly where it is. The question for investors is whether Cognibotics has correctly calibrated its own path to market.
Sources
- [Cognibotics] Cognibotics - Advanced motion made simple | https://www.cognibotics.com/en
- [Cognibotics, 2026] Series A funding announcement | https://www.cognibotics.com/en
- [Mynewsdesk] Order for CogniCal robot calibration technology | https://www.mynewsdesk.com
- [RobotToday] CogniCal product description | https://www.robottoday.com
- [The Robot Report, 2025] HKM1800 design details | https://www.therobotreport.com
- [Robotics 24/7] HKM1800 pick rate claim | https://www.robotics247.com
- [Fairpoint Capital] Newton ApS funding round | https://www.fairpointcapital.com
- [Ideon Science Park, August 2025] The Future Of Fast, Smooth, Elegant Robotics Starts In Lund | https://www.ideon.se/news/the-future-of-fast-smooth-elegant-robotics-starts-in-lund/
- [Robotics and Automation News, 2026] CAISA project funding | https://roboticsandautomationnews.com
- [IEEE Xplore] Klas Nilsson academic profile | https://ieeexplore.ieee.org
- [Alessandro Genua - Scuola Superiore Sant'Anna | LinkedIn] Klas Nilsson experience citation | https://www.linkedin.com
- [PERPLEXITY SONAR PRO BRIEF] Company profile and employee range | (Source summary)
- [Tracxn, 2025] Total funding figure | https://tracxn.com/d/companies/cognibotics