smartWhales Satellite-Based Detection System (SBDS) Leverages AI and Space-Based Imagery for Whale Detection, Driving Critical Conversation Outcomes


In collaboration with Hatfield Consultants (Hatfield), Dr. Kim Davies from the University of New Brunswick, and Dr. Sean Brillant from the Canadian Wildlife Federation, AltaML has published a peer-reviewed paper highlighting the smartWhales Space-Based Detection System (SBDS), aimed at supporting the conservation of the critically endangered North Atlantic Right Whale (NARW). With fewer than 370 of these whales remaining and less than 70 being reproductively active, the smartWhales SBDS expands knowledge and significantly advances efforts to conserve and protect the NARW in Canadian waters.

Since 2021, the team has been evaluating the effectiveness of current Earth observation satellite and artificial intelligence (AI) technologies for detecting NARWs to support current management measures and marine science.

The methodology involved direct satellite image tasking over two NARW foraging aggregation areas and using a convolutional neural network to automate whale detection. The study leveraged the largest dataset of satellite imagery of NARWs to date, testing the feasibility of automated detection.

The goal was to develop and assess an end-to-end, semi-automated process to obtain opportunistic detections of critically endangered NARW and their close relatives, southern right whales (E. australis) from very-high-resolution satellite imagery in offshore-temperate, coastal-temperate, and coastal-tropical whale habitats.


“Today’s tools may not be perfect, but they offer real benefits that can enhance policy implementation and balance industry, economy, and the environment,” explained Vinoth Babu, Gen AI Practice Lead at AltaML. “We’re developing tech to cut satellite costs and improve species identification, proving that we, as a society, can adopt and refine these tools for the greater good.”

The findings demonstrated significant technical promise in large-scale detections of whales in Canadian waters, achieving heightened recall rates (greater than 0.73) using a machine learning model trained on a diverse array of images. The study also highlights current challenges requiring further research and improvements to the SBDS, such as the variability in precision between NARW feeding grounds and coastal tropical right whale habitats.  
 

A table showing machine learning detection model performance statistics on all validation images collected in North Atlantic and southern right whale habitats.


“We’ve made remarkable advancements in demonstrating the combined use of satellite Earth observation and machine learning technologies for the detection of right whales in Canadian waters, such as the Gulf of St. Lawrence. The team is eager to tackle the challenges identified in the study so that the SBDS becomes an operational tool to help protect and conserve the North Atlantic Right Whale,” said Olivier Tsui, Partner and Director of Remote Sensing and Geomatics at Hatfield Consultants.

Future Directions 

Interest in space-based detection of large whales is increasing due to advances in technology. Research indicates potential for satellite imagery in marine conservation, particularly in coastal areas, though challenges remain in northern habitats. As high-resolution satellites and machine learning improve, so does the accuracy of detection.

According to Babu, space-based technology offers two key benefits: it provides a better understanding of animal behaviour and migratory patterns, and it enhances conservation efforts as systems mature. 

“To support conservation, we need more precise species identification. That’s where GenWhales, a secondary project aimed at addressing challenges in marine mammal conservation and disaster prevention, comes into play. The team is researching synthetic images to enhance detection accuracy and reduce data acquisition costs, further advancing species identification,” said Babu.

The smartWhales Space-Based Detection System (SBDS) is undertaken with financial support from the Canadian Space Agency, Fisheries and Oceans Canada, and Transport Canada through the smartWhales Initiative. 

The complete scientific publication is accessible via the following link. Further information regarding the Canadian Space Agency’s smartWhales Initiative and the Space-Based Detection System (SBDS) is available on the storymap.


About AltaML

AltaML is a leader in developing customized applied artificial intelligence (AI) solutions for private and public sector organizations. With a deep commitment to Responsible AI (RAI), AltaML leverages vertical AI and agentic solutions to seamlessly integrate AI into traditional industries, transforming operations, unlocking new revenue streams, and driving significant value for clients.

About Hatfield

Hatfield is a multidisciplinary, Canadian owner-operated company that provides consulting and applied research services in environmental, social, and geomatics sciences to public and private-sector clients. We are technical leaders in the use of remote sensing and geospatial technology to acquire, manage, and analyze data to deliver projects and strategic insights for environmental monitoring and management. We regularly support international space agencies, international finance institutions, and major energy and mining companies with scientific and strategic consulting and data analytics. Hatfield has office locations across western Canada, including North Vancouver, New Westminster, Terrace, Fort St. John, Calgary, and Fort McMurray, and international offices in Indonesia and Botswana.

Media Contacts: 

Alicia Bedard
Director, Marketing and Communications
AltaML
alicia.bedard@altaml.com

Olivier Tsui
Partner and Director of Remote Sensing and Geomatics
Hatfield
otsui@hatfieldgroup.com