Jibo

Project Challenge

With the production release date looming and thousands of robots coming off the manufacturing line, Jibo needed to develop test software that would assess the health of returned robots that were received at their service and remanufacturing center. Their current process required a trained engineer to evaluate every unit returned, which was costly and time-consuming. This manual process also prevented Jibo from quickly identifying and addressing underlying product issues. Implementing a more efficient system was critical, but Jibo lacked the internal resources to tackle the project.

Project Solution

Jibo turned to Geisel Software to develop software that could control the robot, run through a series of tests, provide the operator with feedback and log the results. A team of Geisel developers created a full Python app, which ran on a Raspberry Pi computer installed in the factory, that could remotely test each robot. They also created the database structure and a clear, simple GUI. Geisel’s custom solution could interface with the available Jibo APIs for testing while ensuring the devices remained secure.

Results

Geisel worked with the Jibo team and the service and remanufacturing center to deploy the new testing software. With the user-friendly app, factory workers could quickly conduct a series of more than 20 tests that assessed every bit of a Jibo’s functionality, such as motor, touchscreen, sensors and Wi-Fi radio, and be instantly notified if a unit was malfunctioning. Geisel’s solution reduced total testing time on an individual robot from one hour to less than five minutes – a 92% time savings! The new app allowed the company to efficiently determine and track why the units were returned which delivered faster service, increased customer retention and reduced losses.