In a wooded search trial involving 32 dog–handler teams, performance was associated with weather, experience, management, speed, ground coverage and canine arousal. Differences between handler and instructor ratings indicated possible bias.
Published on IPP:
AIResQ provides 9,788 high-resolution thermal images and a drone benchmark covering varied terrain, seasons and atypical casualty poses. Models trained on it generally improved benchmark detection, but human operators remained more accurate overall.
Published on IPP:
Kelly argues that search assurance should examine the reasoning behind decisions—not only plans, coverage and completed tasks—and proposes recording assumptions, uncertainty, confidence and adaptation triggers at key decision points.
Published on IPP:
Three ways into the articles. Pick one to see only that kind of content.