The engineers, educators, designers and entrepreneurs at Robolink have some Kickstarter experience under their belts and are back with a new robotics tool. After reading the McKinsey report that says by 2030 seventy percent of companies may have adopted artificial intelligence, the team was concerned that no easy resources were available to learn AI. They built Zumi as an educational tool for teaching artificial intelligence and self-driving automotive technology. Robolink is running this Kickstarter campaign to fund their first round of production parts after doing prototype and play testing in California and Korea.
Zumi is built from a Raspberry Pi Zero and is compatible with Arduino as well. The system contains a gyrometer, accelerometer, two DC motors for propulsion, six IR sensors, a buzzer, and a Pi Cam. Using Zumi can teach robotics, mapping and navigation, machine vision, self-driving decision making, machine learning, and Python. TensorFlow and Open CV work with the system and are currently used by self-driving car engineers. Examples on the campaign page show the bot calculating optimal distances to a point, identifying objects and making stop / go decisions, gesture recognition, and face recognition. Beginning programmers can use the integrated development environment already in place and available as a web-based tool, while more advanced programmers can use TensorFlow or Keras to write code.
Zumi looks like a nice addition to the already crowded field of robotics and programming education tools, emphasizing the artificial intelligence and self-driving car angles will help to differentiate them. This is their third campaign and the previous two products are already in production with this campaign raising ninety nine percent of the funding goal as of this writing. Zumi also won a CES 2019 Innovation Award last week at the Consumer Electronics Show in the Robotics and Drones category. The Kickstarter campaign ends on March 9, 2019.