Process/Brainstorming

The overarching issue that we wanted to explore with this IoT device was whether a device could differentiate between aggression and affection in different contexts. 

We began by exploring the breadth of possible human-robot interactions in where there was room for error and problematic results. We then typified these potential problematic results:

We used a 6-3-5 brainstorming session to explore ideas stemming from these paradigms, and we were subsequently able to narrow down our focus to the regimes of "Affection misunderstood as aggression" and "Errors of food automation".

After exploring affectionate gestures that could be misconstrued as aggressive ones based on cultural norms, we determined that there can be a huge difference between gestures displayed towards friends versus those towards strangers. For example, a friend running towards you for a hug might move at the same speed as a stranger trying to attack/rob you. So, we imagined a device that would interpret this movement, but not necessarily be able to distinguish between the situations. Additionally, we assessed the physical signs of distress that occur in a victim before attack. After finding that both heart rate and blood pressure spike in this situation, we imagined the same device having the capability to track these stats. 

This brainstorming session raised several questions. Would the device register two bodies accelerating towards each other as an attack? Would someone slipping a surprise note into a friend's backpack alert the device of a pickpocket? Giving a presentation in front of a lot of people might trigger nervousness and increased heart rate for some people, but would the device interpret this setting as its wearer being threatened by a large group of people? Since these situations will give similar input to the Ghost Pepper Spray device, this IoT product would be dangerously flawed. We show in our video how two friends excitedly reuniting can trigger the device, along with the possible initiation of the pepper spray during an interview.