Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - DINGO SME-1 (Digital avalanche safety - Rethink the off-piste and backcountry skiing experience)

Teaser

Off-piste skiing has grown tremendously in popularity, and 1 in 3 skiers and snowboarders now venture away from prepared slopes. This increases exposure to steep and avalanche-prone terrain, so the number of skiers who have been harmed by avalanches is steadily growing...

Summary

Off-piste skiing has grown tremendously in popularity, and 1 in 3 skiers and snowboarders now venture away from prepared slopes. This increases exposure to steep and avalanche-prone terrain, so the number of skiers who have been harmed by avalanches is steadily growing. Current avalanche products are designed for rescue rather than prevention. Skiers must rely on their own interpretation of the snow conditions by digging snow profiles and checking surrounding topology and snow build up to avoid risky areas. Digging in the snow is time-consuming and the interpretation requires significant expertise that many skiers do not have.

Knowledge of avalanche danger is essential for safe skiing, especially in remote, high mountain areas. Regional forecasting, topographic maps and weather reports offer avalanche risk indicators, but these are based on extrapolations from few data points and do not offer accurate location-specific information. Travel in avalanche-prone terrain is not limited to skiers. Snowboarders, snowmobilers, climbers, militaries, industries (hydropower, construction, etc.), and researchers all travel in avalanche-prone areas. Creating a device for better assessment of snow conditions and danger prediction can help keep all winter travelers safe.
In addition to safe travel, the data collected from these devices is valuable for ski resorts, the hydropower industry, and research organizations.

The main objective of this project is to design a sensor device, small and light enough that it can be mounted on a ski. It is capable of detecting the presence of a weak snow layer with a resolution of a few centimeters where the skier is located and it informs the skier in real-time.
Secondary objectives are:
- recognize other types of snow layers for research purposes
- combine slope inclination, snow layer information and GPS to build maps that can be used for trip planning
- combine this data with other sensor data (gyro) and source of information (weather, avalanche forecasts) to estimate localized avalanche danger
- work with ski manufacturers and the data we generate to evaluate improvements to the ski design that could minimize the risk of triggering an avalanche

Work performed

Please find an update on the 4 work packages listed in our application.

1. Manufacture and test 20 prototypes of Ã…snes skis with DINGO antennas.
The 20 Ã…snes skis with our custom designed radar antenna embedded inside, were produced at a ski factory in Czechia. We have also produced skis with two new brands, SaltSkis and EVI, to test on a variety of ski types and manufacturing processes. We have tested various glues and checked the bounding between the antenna and the wooden ski core. We have also evaluated installing inserts for screws in the ski to have a more robust attachment point for our sensor device. We have performed bend and stress tests on the skis, both of which were successful. We are now testing the skis in the mountains in real conditions to see how the antenna and skis hold up in extreme winter terrain.

2. Continue data acquisition and testing in real snow conditions. Finish construction of radar/sensor hardware device.
In 6 months, we have made significant progress on the design of the ski mounted radar. The starting point was to secure the right suppliers for electronic design, firmware, electronic manufacturing and production of plastic parts. We have selected Norwegian companies to make collaboration easier during the prototyping phase of the project. From August to October, we developed the design in collaboration with our partners. We launched the production of 15 electronic board prototypes and 15 plastic housings (the box around the battery and the electronic) in October. We received the plastic housing on the 12th of November and the electronic boards in early December. We used the months of December and January to program the boards, fix issues and install them into the skis. We are currently testing the fully assembled units on skis with our technology manufactured inside. These prototypes are being used by our team as well as expert backcountry guides. The goal of these tests is threefold: mechanical testing, data collection, and user experience. Mechanically we are testing that all parts fit together, do not compromise the integrity of the ski and remain robust and watertight. For data collection, we are testing and verifying firmware functionality, as well as data quality. It is important to verify that our design and the data collected are of high enough quality to process and extract the necessary data. For the user experience, we are obtaining feedback regarding the device design and communication mechanism, as well as the type of information users find necessary.

3. Begin developing the DINGO software platform for collection and sharing of data.
The high-level architecture of the software platform to store the data has been defined and built. The platform solution uses Google Cloud and a Firebase database. Our hardware devices will be connected using an IoT (Internet of Things) infrastructure, which will enable each device to receive firmware updates, as well as send its data when an internet connection is available. We aim to test the data upload in May 2019. We are continuing development on the software platform so it can hold and process all the data collected from the sensor devices.

4. Start phase 2 of an on-going R&D project developing advanced layer detection based on machine learning.
We have submitted an application with the Norwegian research council for “User-driven Research-based Innovation (BIA) “. In September we worked to secure strategic partnerships with research institutes and other businesses interested in our technology and data. Our research partners are SINTEF (www.sintef.no) and NGI, Norwegian Geotechnical Institute (www.ngi.no). SINTEF will focus on data quality (signal processing, antenna design, shielding from interferences, etc.) and NGI will focus on the machine learning algorithm, data labeling, and field data acquisition. For the business partners, we have our existing partnership with the Norwegian ski manufacturer, Åsnes, and we have settled

Final results

By the end of this project SME-1, we will have tested our prototypes in real conditions and acquired snow data to start training our machine learning algorithms. The engineering effort to make our sensor device smaller, cheaper and more attractive for customers will have started. We have also sold prototypes to the leading backcountry guide network in Norway, DNT. Finally, we started the research collaboration with our 4 partners to benefit from their expertise and received funding from the Norwegian Research Council to support this collaboration.

The long term impact of this project will lead to a more thorough understanding of the snowpack, local snow conditions, and avalanche risk. The insights generated from these devices will bring value to many markets outside of skiing, including hydropower companies, ski resorts, weather forecasting systems, militaries, roads and railways, construction, research, climate change, etc. By combining detailed localized snow information with different data sources (weather, topological, aspect, historical avalanche data, avalanche forecasts) a more accurate assessment of localized avalanche risk will be generated. This information will help increase safety for all backcountry travelers and ultimately help save lives.

Website & more info

More info: http://www.snow.ski.