dew•E - Gardening in AR with robot tasking.

Segway Loomo monitoring a test planter bed

Inspiration:

We are inspired by the very feasible prospect of combining spatial computing and robotics to do something as meaningful as growing food...

where there wasn't before,

by the average person - like us.

What it does:

dew•E solves for the creation and maintenance of a food-producing garden. dew•E provides the path of least resistance for people and communities to grow their own food.

dew•E consists of a user experience empowered by AR + cloud analytics + robotic automation.

The AR app assists the user plan, plant and predict their garden harvests with the robot tasked to provide a per plant status update.

How I intend to built it:

We plan to use the Unity to develop a way to survey a plot, recognize individual plants and task our robot to perform periodic tasks for each plant.
We are going to design an intuitive, mixed reality interface that allows a user to see the status of each plant, updated from the robot.
We intend to use our Segway Loomo robot to serve as our semi-autonomous robot for this hackathon.
We are using Tensorflow object recognition libraries to serve as navigation for our robot.
Our cloud stack enables the user to interact with the robot and send data back and forth.

Challenges that we are running into:

Developing a recognition system for plants.
Getting the robot to navigate its surroundings effectively.
Making the user experience easy to interact with.

Accomplishments

This same core team has previously won an VR/AR AT&T Hackathon in Bellevue WA in Feb. 2017. See this link for winning project details: [https://www.thegrid.ai/carrottop/]