PRESS RELEASE: Huawei and BJFF won a GSMA GLOMO award at MWC Barcelona 2024 for “Outstanding Mobile Contribution to the UN SDGs” with an AI-powered solution designed to prevent the extinction of Norway’s wild Atlantic salmon.

“We are very honored to accept the GSMA GLOMO for our TECH4ALL project, and special thanks to our partner BJFF for their commitment and support,” said Lu Yong, Huawei Senior Vice President and President of the European Region for Huawei. “The salmon project shows the great potential of technology to conserve nature while driving the sustainable development of local communities. Huawei will continue to use innovative technologies to build a greener world.”

Launched in 2021 under Huawei’s TECH4ALL initiative and developed with local partners BJFF (a fishing and conservation group), Simula Consulting, and Troll Systems, the solution was developed over three years. It is designed to respond to an invasive species, the humpback salmon, which threatens local Atlantic salmon in Norway’s rivers.

The solution, the world’s first of its type, filters out humpback salmon from the river system, preventing them from proceeding upstream to spawn.

“We’re immensely proud to collaborate with Huawei and partners on this ground-breaking project. The cutting-edge technology solution we have in Berlevåg is a successful recipe for stopping the humpback salmon, and it is inspiring for us to gain global recognition at MWC Barcelona 2024,” said Geir Kristiansen, Chairman of BJFF. “This project not only protects our rivers and wild Atlantic salmon, but also showcases how innovation can tackle environmental challenges worldwide.” Berlevåg is the site in Norway where the technology is deployed and the home of BJFF.

The GLOMO judging panel described the project as an “excellent and scalable idea – the application of self-powered AI and mobile tech to accurately solve a problem, with great execution and impressive results.”

How the solution works

Fish are directed to swim through a river-spanning tunnel installed with an underwater camera and an intelligent algorithm-powered gate.

The algorithm identifies different fish types, recognizing, for example, the distinctive hump of the male humpback salmon and tail patterning on the female. With an identification accuracy rate exceeding 99%, the system instructs the gate to remain closed to filter the invasive species into a holding tank, but opens to let other fish swim upstream.

Between June and September 2023, the latest the salmon spawning season in a biennial cycle, pilot projects in Storleva River and Kongsfjord River successfully filtered out more than 6,000 humpback salmon.

The alternative method, by contrast, requires volunteers to stand in the river, identify humpback salmon by sight, and remove them by hand. Not only is this method labor- and time-intensive, many fish are accidentally injured or killed in the process. In comparison, the automated solution causes no harm to fish and reduces manual workloads by around 90%.