Signal detection theory exemplifies not only how accurate we are at picking up on a stimuli, but also how well we excel at doing so given the individual circumstances. What signal detection theory describes is how each situation can be unique – in terms of severity in misses and false alarms as well as in levels of capability in actually picking up the appropriate signal.
Humans aren’t perfect at reading in our environments. We’re far from it. So when thinking of how automation can extend human capabilities to read in signals, I thought of how we have turned to other alternatives outside of machines to enhance our detection of stimuli. Particularly, my mind went to our use of explosive sniffing dogs.
Dogs make for the perfect candidate given such a task. They are loyal and trainable, and have the smelling hypersensitivity that we lack. We can, and have, catered the training of dogs to contribute both domestically and internationally to our protection in the most stressful of situations.
So do we need automation in this domain when we have dogs? It’s certainly worth investigating. In the last couple weeks, multiple news outlets have reported the neglect and mistreatment of these dogs overseas in Jordanian kennels.Shocking images reveal these highly trained and skilled dogs starving of death and heat stroke, and living in shockingly destitute conditions. Neglect is abundant, as the U.S. continues to ship more of these dogs overseas.
While dogs may seem like a strong alternative to humans to sniff explosives, I argue automation should be our next frontier in this domain. If we aren’t willing to take care of our dogs, we have no business using them. Can we mimic the hunting instincts and smelling abilities in machines? I hope so, it’s at the very least worth exploring.