The KOKO 2 project, formally titled “Collision Detection / Collision Avoidance for Single‑Pilot Cockpit and Unmanned Cargo,” was carried out from 2020 to 2023 under the German Ministry of Defence’s research programme (project code 20V1707A). It built on the earlier KOKO effort and aimed to extend a proven detect‑and‑avoid (DAA) system from business‑jet and commercial‑aircraft support to crew‑reduced and fully unmanned cargo aircraft. The consortium was led by Hensoldt Sensors GmbH, with the Detect and Avoid team in Ulm at the core of the effort. Additional technical support came from Hensoldt’s Multi‑Sensor Datafusion and Avionics Systems groups in Immenstadt, and the project leveraged radar and electro‑optic sensor hardware that had been developed for the ProSA‑n programme. The consortium also included several other German research institutes and industry partners, all contributing to the design, simulation, and validation phases.
The technical work was organised around four main objectives: (1) the creation of an operational concept for crew‑reduced or unmanned transport aircraft; (2) the definition and integration of a wide‑area radar system and a complementary electro‑optical sensor for non‑cooperative, redundant detection; (3) the development of a sensor‑data fusion architecture; and (4) the implementation of a collision‑avoidance logic. A comprehensive simulation environment was built to support these tasks, comprising model‑based systems engineering (MBSE) simulations, pilot‑in‑the‑loop scenarios, and flight‑test‑data‑based runs. The simulation framework enabled rapid iteration of the radar modes, including a dedicated DAA mode, weather‑detection and ground‑mapping functions, and the evaluation of gust and turbulence detection algorithms.
The radar subsystem was designed to meet the minimum operational performance standards set by RTCA DO‑365A for DAA and DO‑220A for airborne weather radar. It incorporated a dual‑mode architecture that could switch between collision‑avoidance and weather‑monitoring functions. The electro‑optical sensor provided a non‑cooperative, dissimilar sensor stream that enhanced detection reliability in cluttered environments. Data fusion combined the radar and electro‑optical inputs to produce a single, high‑confidence intruder track. The collision‑avoidance logic then assessed threat levels, calculated avoidance maneuvers, and issued alerts to the pilot or autonomous flight control system.
During the two flight‑test phases, the demonstrator achieved detection of intruders at ranges exceeding several kilometres and maintained a false‑alarm rate below the thresholds defined in the RTCA standards. Weather‑related features such as gust and turbulence detection were validated against ground truth, confirming the radar’s ability to identify hazardous atmospheric conditions. The system’s performance was documented in a success‑control report that accompanies the final project report, confirming compliance with the agreed‑upon metrics.
The project’s collaborative structure ensured that each partner’s expertise was leveraged efficiently. Hensoldt’s core team handled system definition, simulation, and integration, while the Immenstadt groups focused on data‑fusion algorithms and avionics interfaces. The ProSA‑n partnership supplied proven radar hardware and test facilities, accelerating development. The consortium’s joint effort produced a technology demonstrator that, although not yet certified, satisfies the current European Technical Standard Orders and is positioned for future certification pathways. The KOKO 2 outcome demonstrates that a single‑sensor‑redundant, radar‑based DAA system can support both crew‑reduced and fully unmanned cargo aircraft, paving the way for safer integration of unmanned transport into controlled airspace.