Spring 2018 Winning Project: EVE
The first care companion designed to bring compassion and comfort to the experience of anxiously waiting for news in an emergency room, while helping doctors and nurses maintain focus on treating patients.
Although we designed EVE for everyone, what really interested us in creating a compassionate ER companion was the empathy and passion we have for this use case. Everyone on our team has seen family and friends deal with the pain, confusion, and frustration of navigating this experience on their own.
To make certain our solution would make a positive impact, we began by searching for the most acute pain points of a user’s experience when they find themselves unexpectedly sitting in an emergency room, waiting to hear about a loved one. We found that people were often left to wait for hours on end with little information on the status of the patient. They experienced elevated levels of stress, anxiety, and forgot about their own personal care while focusing all of their energy on the well-being of a person they couldn’t help at that moment.
With AI’s ability to quickly mine insights from vast amounts of data and communicate responses with human-like nuance, we knew this technology could change this experience for the better.
EVE is a service we see hospitals subscribing to and providing for free to patients and guests. Service bots have proven to dramatically reduce resolution time and the cost of offering personalized support, all while increasing customer satisfaction. This leads us to believe that EVE would provide incentivizing value to hospitals.
While EVE will never replace the words of a doctor or even a hug from a stranger, it can help hospitals take better care of everyone who walks through their doors.
EVE is an AI presence that has been created to improve humans lives, not replace them.
Proactive data privacy
The moment a visiting friend or family member checks in with the front desk, they are given the option to have EVE as a support and care assistant by giving it access to personal apps and information. They may have even set this up ahead of time through their doctor. Users have complete control over the information they are comfortable sharing.
They might let EVE see everything — medical IDs, social media, calendars, text, email, GPS, insurance, and banking — so it can provide the most personalized and automated experience. Or they may only allow access to their Facebook and contacts.
When fewer permissions are granted, EVE offers help when moments arise and lets the user decide if they want to share personal information as needed. For example,
EVE may ask “Do you have kids?” based on their social photos and the fact that it is 2:30 pm — close to the time school lets out. If the answer is yes, EVE can offer to call the babysitter, or order them food.
To help users take better care of themselves, EVE is always available to talk through anxious moments, provide patient updates, give details about symptoms and treatments, or merely be a guide to the nearest restroom. When it is not actively communicating, EVE is working in the background to anticipate a user’s needs and next steps.
Mapping compassion
From the start, we knew creating a solution for this use case was going to be a huge challenge. We wanted EVE to provide care and communicate compassion between everyone involved in hospital experiences — not only patients, friends and family, but doctors and nurses too.
To design an experience capable of this level of service, we had to map the components of compassionate human communication.
We started by identifying the roles EVE plays across our user groups:
Next, we evaluated the practices and ethics a system like EVE would need to follow in each of these unique roles. Our solution was to create a Hierarchy of Naturalization that established a series of filters for EVE to run all potential outputs through:
Priority
Mental state
Literacy
Environment
Timing and choreography
Privacy
Security
Demographic
EVE will be continually crafting and considering an array of potential responses and the appropriate way to express them.
By applying confidence scores to the “dials” within each of these filters, EVE will be able to know from day one how to choose the response that’s most likely to have a useful outcome. Over time, with more interactions, EVE will continuously be adjusting scores and improving based on direct user feedback.
Accessibility
To be genuinely compassionate, EVE needed to be accessible to people from all backgrounds while in a highly emotional situation.
Our solution for this was freeing EVE from being tied to one format — an app, a roaming robot, or a virtual assistant. EVE is an experience that’s available for every level of accessibility.
Are you talking to a young child or someone with hearing loss? Use EVE as an avatar that can be called upon any digital device to communicate with expression, sign language or closed captions.
Helping someone wrought with anxiety and panic in the middle of the night? EVE can become the soothing connection between patient and loved one by emanating a soft light through an iPhone that pulses with the patient’s breathing.
Visually impaired? EVE becomes a comforting voice on the other end of the phone. Busy driving or running an errand? EVE reaches out through a text message.
That is what EVE is all about — considering the health and well-being of everyone in a hospital and working to improve each experience with support, information, and compassion.
The EVE team
Joel Swiatek, Computer Science
Grayson Rosato, Theater, Design and Management Strategies
Anumeha Sinha, User Interaction Design
Veena Suthendran, Business Honors & Management Information Systems
Gabriela Sanders, Design & Human Centered Interaction
Abishek Ramani, Mechanical Engineering
