Insight
4 min read
Imagine a drone that needs to fly a search pattern over an area with no mobile coverage, at night, with no one sitting behind a laptop. That is precisely the situation for which Spectro-AI, based in Enschede, builds its systems. The company develops AI software and matching hardware that allows autonomous systems, drones first and foremost, to run entirely locally: no cloud connection, no fixed ground station, no operator directing every move.
Founder Hamed Mehdipoor, an engineer by background, started in 2018 by analysing and modelling spectral drone data for forestry and agriculture. Since then, the company has grown into a supplier of integrated hardware and software solutions for autonomous inspections across multiple sectors. The team numbers seven people and works from the Marssteden business park in Enschede.
Spectro-AI positions itself at the intersection of deep tech, AI and robotics, a segment where the gap between the laboratory and practical application is significant. The company deliberately opts for that practical side: systems that work in the field, even when connectivity drops.
Spectro-AI's product portfolio consists of three components that together form a fully autonomous system. At the centre is SAI-HUB, the operational software that combines flight planning, mission execution and data processing. Through SAI-HUB, operators can create missions, deploy drones and receive automatically generated reports containing images, video and object detections. The software supports rapid mission creation and is designed for situations where time is critical.
The Brain-Box is the local processing unit that provides SAI-HUB with computing power. The unit stores detection algorithms and processes image data at up to 60 frames per second. Because processing happens locally, data does not leave the site. The Brain-Box has a built-in SIM card for SMS alerts, including in areas where connectivity is not always stable.
The third product is the DockWatch Trailera mobile drone dock mounted on a trailer. The trailer serves as a self-contained operational hub from which a drone takes off, carries out its mission, returns and recharges, without any fixed ground infrastructure. The trailer is intended for rapid deployment in the field, in rugged or remote environments. In April 2026, Spectro-AI presented the DockWatch Trailer Dual at the eRIC2026 Expo.
Spectro-AI's roots lie in the analysis of spectral drone data for agricultural and forestry applications. That origin is still evident in the emphasis on automated data processing and detection algorithms, but the scope of application has since broadened. The company now serves a range of sectors, from industry and infrastructure to public safety.
One concrete example is the collaboration with SOS Drone Team NL, an organisation that deploys drones in missing-persons cases. The two parties conducted a joint exercise with the DockWatch Trailer Dual and SAI-HUB DD, during which the system flew a search pattern. That type of deployment sets high demands: the system must operate reliably in unfamiliar terrain, outside office hours and without constant supervision.
In December 2025, Queen Máxima initiated a mission on SAI-HUB DD during a visit to Drone Lab Amsterdam. That visit drew attention to the Dutch drone sector, of which Spectro-AI is a part. Earlier in 2026, the company announced a partnership with CopterPro, a party with expertise in aerial systems and algorithm development.
Spectro-AI has seven employees. For a company delivering integrated hardware and software for autonomous systems across multiple sectors, that is a small team. Available sources are inconsistent on external funding: the DutchStartup database lists a Series B round of €19.3 million in January 2022, while verified web information indicates that Spectro-AI has not raised any funding rounds. The two sources diverge, and the actual funding situation cannot be established with certainty on the basis of this material.
What is clear is that the company is active in a rapidly growing market. Autonomous drones for industrial inspection, surveillance and public safety are attracting increasing interest from both commercial and government parties. Demand for systems that work off-grid and without permanent operator control stems partly from data-storage regulations and partly from operational reality: a stable internet connection cannot be taken for granted in the field.
Within the Dutch AI and robotics scene, Spectro-AI is an example of a company that takes a specific technical problem, local AI processing for autonomous systems, as its starting point and from there serves multiple application domains. Through the University of Twente and its surrounding ecosystem, Enschede is a fertile environment for this type of deep-tech company. Whether the growth path Spectro-AI has in mind is achievable with a team of seven remains to be seen in the years ahead.
Amsterdam-based ThreatFabric helps banks and financial institutions detect fraud, scams and malware across web and mobile channels. The company, founded in 2015, now protects more than 60 million online banking customers worldwide.
The Dutch biotech and life sciences sector counts a growing number of AI-driven startups in 2026, while at the same time seeing fewer new company formations and a concentration of capital among a handful of players. Thirteen companies provide a picture of where the sector stands.
Rotterdam-based Odd.Bot is developing autonomous weeding robots for arable farming. Their robot, The Maverick, runs on electricity, uses no chemicals, and can cover up to two hectares per day on crops such as carrots, onions and beetroot.
