About Ozense

Ozense is a pre-founding research and development project. No project-related company has been incorporated; "Ozense Radar Systems" is a working name. The legal form will be settled only at a point permitted under the funding rules.

The goal is a pilot-capable proof of concept: a cost-efficient, locally operating system that detects and tracks small non-cooperative drones and conservatively distinguishes them from birds, other airborne objects, clutter and unknown cases.

What Ozense is not today: a supplier with available products, a systems house with reference installations, or a counter-drone effector. There is no released product price, no series hardware and no completed field validation.

Hakan Özen

M.Sc. Aerospace Engineering (RWTH Aachen University) and B.Sc. Mechanical Engineering (Paderborn University). Professional background as a System Test Engineer at BITES Defence and Space, as an IT consultant at NTT Data Germany, and in research-adjacent simulation work at the WZL of RWTH Aachen.

Responsible within the project for radar and signal processing, tracking and classification, edge software and technical architecture, and measurement methodology and ground truth. The existing Ozense radar, tracking and validation stack was built independently.

A three-person team is envisaged for the planned funding phase, adding market and sales as well as hardware and field build-up. These roles are planned and not yet finally settled; the people concerned are not currently employees.

The project is deliberately described as technically asymmetric: radar signal processing, machine learning, edge software and systems integration depend heavily on one person. An external radar, RF or measurement partner is therefore explicitly foreseen for range calibration, antenna and RCS questions, and independent technical credibility.

The ambition of this project is less to produce good numbers than defensible ones. A few fixed rules follow from that:

• Every technical performance claim needs a reproducible evidence path and a stated limit.
• Development, validation and independent holdout data stay separate.
• Evaluation thresholds are pre-registered and not readjusted after a holdout has been seen.
• Market and price assumptions stay flagged as hypotheses until interviews, quotes or contracts support them.
• Where marketing text and reproducible technical evidence conflict, the evidence wins.

This website is derived from those rules. Where it states a limit, that is not understatement but the actual status. The full development status is documented here.

The intended funding route is the German EXIST start-up grant, with Paderborn University and garage33 as the university environment. The funding phase is meant to enable the step from the existing radar proof of principle to a defensible radar classification proof of concept.

While that clarification is under way, self-imposed constraints apply: no project-related company before the permitted point in time, no paid customer-specific contract development, and research pilots strictly unpaid and non-exclusive.