Inspiration
Coronavirus.
What it does
A simple questionnaire is presented to citizens. Patients are immediately told of the chances that they have Coronavirus and often are given advice on how to further avoid the virus. When a patient responds to the questionnaire with worrisome symptoms, the patient may be forced or advised to go to a local hospital, depending on the province. When the machine learning model reaches a certain confidence threshold of a severe Coronavirus infection, Celeste immediately kills the patient.
The algorithm behind Celeste's judgement is a novel, state-of-the-art machine learning model able to diagnose patients with an estimated 87.9% accuracy in tens of thousands of simulated trial runs based on global data reported up to 14:02 UTC, Sunday, February 16, 2020.
With six legs and eighteen motors, Celeste can effortlessly climb up stairs and is ideal for the massive apartment complexes in China that lack wheelchair accessible ramps for wheeled robots. Celeste's arms are able to forcefully knock on doors and, thanks to an on-board internal amplifier and a vacuum-tube speaker, she is also able to verbally communicate with patients and realistically reproduce Chinese phonemes with commanding accuracy.
How I built it
3D printers and zip ties
Challenges I ran into
Celeste herself became infected. As a result, her motor controller burnt out.
Accomplishments that I'm proud of
Being able to access huge amounts of real data and attempt to create actionable intelligence for a truly worrisome problem
What I learned
What's next for Celeste the robot tackles Coronavirus
Celeste takes over the world
Log in or sign up for Devpost to join the conversation.