Current State of Fish Monitoring
Our experience tracking tens of millions of adult salmon throughout the Columbia River Basin and with migratory species throughout North America has provided key insights into the logistical challenges associated with fish enumeration efforts. Technicians are often hired to observe fish passage at ladders in real-time or on recorded video collected through a window beneath the water surface. This method is labor-intensive, relying on observers watching hours of live or recorded video. Projects are frequently located in remote or highly secure areas, creating additional challenges in hiring and coordinating staff on a seasonal basis. Maintaining data quality can be challenging at a minimum, with missing data or errors that can influence management decisions.
Seeing these challenges as an opportunity for improvement, Four Peaks and our partner, Boulder AI, developed BlueFish. BlueFish is a computer vision system comprised of an edge-computing camera and customized software that work together to enhance the collection and distribution of fish enumeration data from remote locations. Our state-of-the-art system provides the ability to monitor fishways around the clock and deliver data over the internet, allowing managers, biologists, and technicians to focus on restoration actions rather than tedious monitoring activities and staffing logistics.
To date, Four Peaks has successfully tested and deployed BlueFish at federal and regional hydroelectric projects across the Pacific Northwest. BlueFish has improved fishery management efforts at these projects by increasing fish count accuracy, expanding counts to hours and regions where human-staffing proves challenging, and streamlining cumbersome data collection and distribution processes.
Specific benefits include:
- Data collection that is two-thirds faster compared to manual live or video counts
- Integrated quality control and early identification of issues
- Lower labor costs and reduced staff coordination time
- Statistically-robust results managed in the cloud and accessible on demand
Using machine learning, the platform also has the capability to add an increasing number of species and distinguishing characteristics to its knowledgebase, further reducing staff time requirements for fish count acquisition. Our team is continuously working on building new features and improving BlueFish with an eye towards a future of completely automated fish counts.
Contact us at email@example.com to discuss how we can help you achieve your fish monitoring goals.
Video produced by Voortex Productions