Emotionally Responsive Microwave
Designing emotionally intelligent conversational interface for home appliances to help the user make healthier choices
The project is a year-long comprehensive research and design of intelligent machine interfaces and languages
including numerous prototypes, research tools, and interviews with users and experts.
Machine's Emotionally Intelligent Traits
Helps you reflect on your actions and does not make decisions for you
Refers to past evidence of a
good choice for making suggestions
Knows when to intervene
and when to keep quiet
Admits failure graciously
and adapts new methods
1. Conversation Starters
Research tools to aid participant’s imagination of future scenarios and
gather design-problem statements pertaining to their experience with objects with futuristic tech-interfaces.
Analysis of the narratives and information from the conversations to distill insights to work further.
2. Insight Distillation
3. Diegetic Prototypes
Prototyping from the design narratives and testing it with the participants to generate deeper insights about people’s preferences and expectations with technological interfaces.
4. Theory Formulation
Building Emotional Intelligence into Machines:
We have a tendency to give human attributes to objects that appear to be autonomous and we form relationships with them. Objects with helpful anthropomorphic behaviors could lead to better and healthier interactions. If objects are programmed to respond to users’ affect, the responsiveness will be perceived as the machine's emotional intelligence.
This perceived emotional intelligence involves a two-stage capacity:
ACQUISITION: its ability to acquire information about the emotions from electronic sensors and learn users behavior patterns from the interactions and external sources of information such as the Internet.
APPLICATION: its ability to act (through appropriate programming) considering the user’s affect based on the acquired knowledge.
5. Hacked Microwave
The first prototype came from a user's narrative originating from Story Engine and the first microwave was prototyped in a week. The microwave detects the food using an image-recognition API. Based on whether the choice is healthy for the user the microwave makes a "happy" or a "sad" sound.
The magnetron (microwave emitter) is disconnected to avoid any accidents.
Only the fan, light and the tray motor runs when it is switched on, giving an impression that it is working. Materials: Microwave + Arduino + Max
Insight from iteration #1
Responses generated by the Microwave might force users to make decisions due to the guilt of "making the microwave sad". Hence, intelligent objects should not make any judgments about the users' choices. the response should only inform the user better so they can make an informed decision. Brainstorming and interviews with Psychologists and participants led to the inference that if the machines could adopt traits from helpful professionals, in this case - a therapist - machines would be much more effective.
Testing new interface with Psychologist Jessica Miller
Building the final prototype.
At Devoxx US, Sanjose, CA
Research Strategy Advisor:
Charlie Cannon (Department Head, Industrial Design, RISD)
Ryan Bardsley (VP, Artificial Intelligence, AiCure)
Joy Ko (AI and Behavior Science) (Professor, RISD)
Peter Snyder (Neurology and Behaviour Science) (Senior VP and Chief Research Officer at Lifespan Health System)
Jessica Miller (Clinical Psychologist, RISD)
David Kim (Program Manager & Creative Technologist, CoWorks Lab, RISD)
Brian Kane (Professor, RISD)
Tanmay Dharmaraj (SharePoint Consultant, RapidCircle, India)
Tim Maly (Professor, RISD)
Paolo Cardini (Professor, RISD)
Claudia Rebola (Professor, RISD)
Khipra Nichols (Professor, RISD)
Chris Novello (Critic, RISD)
Nathan Davis (Mathematician)
Ingrid Fetell Lee (Design Director, IDEO)
Bill Stewart (Director, Design Strategy, Fidelity Labs)
Tom Weis (Professor, RISD)
Scott Geiger (Professor, RISD)
Ayako Takase (Professor, RISD)