Addressing the Challenges of Drone Operations in Icing Conditions

NORCE invited participants in the drone community to an open think tank at Oscarborg to share techniques, procedures, and technical solutions related to drone flying under winter conditions.

NORCE's drone community has many years of experience from flights under winter conditions, both in the Arctic and Antarctic, and Nils Håheim-Saers has previously conducted a vulnerability study describing various challenges related to drone operations in the Arctic. He believes it is important to facilitate knowledge sharing and open conversations to find solutions that work.

– We have invited to this think tank as an open arena for everyone who is interested in and affected by the topic," says Senior Engineer Nils Håheim-Saers at NORCE. As part of our societal mission, we want to contribute to creating meeting places so that especially those who use drones on behalf of all of us have an easier task.

No icing sensor

Speakers at the think tank included Jørgen Lunde Høstvik Ronge (Police Drone Service) and Sigurd Harsheim (Army Maneuver School). Both the Army and the Police use drones in operations to a much greater extent now than before, and the police now have over 100 drone pilots. The lack of sensors means that in many cases, pilots have to make their own assessments of weather conditions. Sometimes these assessments are too coarse, resulting in icing not being detected until it's too late.

Can precise weather forecasts and observations directly support civilian and military drone pilots in the field?

Susanne Foldvik from the Meteorological Institute described several types of weather conditions as challenging. Supercooled droplets (rain) are the most demanding. But problems can also arise where liquid droplets hit a cold hull (inversion), in freezing fog, or in mountain wave clouds that contain a lot of water vapor.

Drone pilots can use various sources for weather data before flying, such as Yr and Værio from the Meteorological Institute. Meteomatics is working on deploying meteodrones at key locations in Norway and will soon launch a tailored service for users in the field, available through a mobile app. However, ground assessments can be difficult. Some drones operate over longer distances, and weather conditions can be completely different a kilometer away from the pilot's location.

Systematic Assessments

Until drone-mounted sensors are available to indicate risk, pilots must "feel the weather," use available weather data, and preferably conduct short test flights.

Training Manager Andreas Nilsen in the Police referred to a report from the Norwegian Defence Research Establishment, which described that the first indication of icing would be visible after two minutes, and after five minutes, icing would be guaranteed given the "right" weather conditions. This "model" is based on research data from the weather service Meteomatics. In line with this, it may therefore be sensible to conduct test flights of varying lengths.

It is also crucial that the pilot has the competence to recognize signals from the drone that may indicate icing. Nilsen described a drone with little power surplus - it becomes slower.

One may also experience an RPM warning (revolutions per minute) indicating that the engine is working "at high speed," the drone loses altitude, and as the descent rate increases, the drone may exhibit atypical movement patterns. Therefore, it is important to have good procedures for safely landing the drone.

Research Initiative

The NORCE project Gondul investigates whether high-resolution weather forecasts can quantify the risks associated with the use of military drones under Arctic weather conditions. The project will specifically study planning processes at the tactical level in the Army. Specifically, the planning and prioritization of military drone systems (NATO Class 1 UAV systems) will be studied with respect to cloud cover and cloud height under weather conditions with the risk of icing and snowfall.