Back to News Hub
⚙️IEEE Spectrum AI
May 26, 2026
Regulation & Policy

Thermal Cameras and AI Help Ships Steer Clear of Gray Whales

Overview

A collaboration of government agencies and scientists launched an AI-based whale detection system in San Francisco Bay on 19 May to reduce deadly ship strikes on gray whales. Developed by WhaleSpotter, the system uses an AI model to spot whales in thermal camera footage from Angel Island, with a human verifying each detection before nearby ships are warned. The Bay has become a dangerous new stop for gray whales, with ship strikes causing many recent deaths. Researchers also plan to use the data to study the whales and add a ferry-mounted camera for fuller coverage.

Key Takeaways

  • The AI-based whale detection system launched on 19 May to help ships avoid gray whales in San Francisco Bay.
  • WhaleSpotter's AI model detects whales in thermal camera footage from Point Blunt on Angel Island, with detections verified by a human.
  • Verified detections trigger an alert to the U.S. Coast Guard's Vessel Traffic Service, which shares the information with nearby ships.
  • Thermal cameras work 24/7, including at night and in fog, and can spot the hot spout a whale exhales even in a pixel or two.
  • A second camera on a San Francisco-to-Vallejo ferry will be added to improve coverage.

Stats & Key Facts

  • #System launched on 19 May
  • #A study published in April estimates an 18 percent mortality rate for gray whales that enter the bay
  • #Seven gray whales have died in the bay so far in 2026
  • #A record 21 gray whales died in and around the bay in 2025
  • #Necropsies showed 40 percent of those 2025 deaths were caused by ship strikes
  • #Over 100 gray whales have detoured into San Francisco Bay since 2018
  • #Gray whales travel between 15,000 and 20,000 kilometers round-trip
  • #The camera covers a cone of water extending about 7 km
Thermal Cameras and AI Help Ships Steer Clear of Gray Whales

How the Detection System Works

The system pairs thermal cameras with an AI model and human verification.

  • A thermal camera at Point Blunt on Angel Island looks down at the bay from a Coast Guard tower.
  • An AI model detects whales in the footage, then a marine-mammal expert verifies the detection.
  • Verified detections trigger an alert to the U.S. Coast Guard's Vessel Traffic Service, which warns ships so they can slow down or reroute.

The technology was initially developed at the Woods Hole Oceanographic Institution in Massachusetts to recognize the distinctive hot spouts whales exhale, even when they appear in only a pixel or two. The camera covers a cone of water extending about 7 km.

Why Gray Whales Are Dying in the Bay

Ship strikes have made the bay deadly for gray whales.

  • A study published in April estimates an 18 percent mortality rate for gray whales that enter the bay.
  • Seven gray whales have died there so far in 2026.
  • In 2025, a record 21 gray whales died in and around the bay.
  • Necropsies showed 40 percent of those deaths were caused by ship strikes.

Douglas McCauley, director of the Benioff Ocean Science Laboratory at UC Santa Barbara, said last year was truly a crisis for gray whales.

A New Stop on a Long Migration

The Bay Area is a recent detour on the gray whale migration.

  • Gray whales make the longest migration of any mammal, traveling 15,000 to 20,000 kilometers round-trip from Alaska to Baja California, Mexico.
  • Historically they made no stops, gorging before departure and fasting on the journey.
  • Since 2018, over 100 gray whales have detoured into San Francisco Bay on their way north.

Scientists are not sure why the whales are stopping but suspect record-low sea ice driven by rapid climate change is decimating the algae that fertilize the Arctic food chain the whales rely on. Without enough food, the whales may stop in San Francisco Bay to feed, sometimes staying over a month.

Why Thermal Cameras Help

Thermal cameras outperform human spotters in key conditions.

  • When whales surface to breathe, they expel a jet of air hotter than its surroundings.
  • Thermal cameras powered by AI can detect these spouts even when they show up in only a pixel or two.
  • The cameras keep watch 24/7, including at night and on foggy afternoons.

Expanding Coverage and Research

The collaboration plans to grow the system and study the whales.

  • A camera will be added on a ferry running between San Francisco and Vallejo to provide fuller coverage.
  • UC Santa Barbara researchers and scientists at the Marine Mammal Center will use the data to study the whales.
  • The data are also available in an online map.

Frequently Asked Questions

When did the whale detection system launch?

The AI-based whale detection system launched on 19 May in San Francisco Bay.

How does the system avoid false alarms?

Detections from the AI model are verified by a human marine-mammal expert before any alert is sent.

How deadly has the bay been for gray whales?

A study published in April estimates an 18 percent mortality rate for whales that enter the bay, and in 2025 a record 21 gray whales died, with necropsies showing 40 percent of those deaths were caused by ship strikes.

Why are gray whales stopping in San Francisco Bay?

Scientists suspect record-low sea ice driven by climate change is reducing the Arctic algae whales rely on, leaving them too hungry to complete their migration without a feeding stop.

How will coverage be expanded?

The collaboration will add a thermal camera on a ferry running between San Francisco and Vallejo to provide a fuller picture of whale activity.

By pairing thermal cameras with AI and human verification, the project aims to warn ships in time to keep gray whales safer in San Francisco Bay.

Continue Learning

Originally published by IEEE Spectrum AI
Read the original

Comments

Sign in to join the conversation