Fresno State builds an autonomous Bulldog Bot with AWS RoboMaker

Fresno State Bulldog Autonomous Bot Team with their Bulldog Bot

During Homecoming 2019, the California State University, Fresno (Fresno State) Hub of Digital Transformation showcased the Bulldog Bot – a robotic take on the school mascot – built with Amazon Web Services (AWS) RoboMaker and AWS artificial intelligence (AI) services. The interactive Bulldog Bot engaged fans across campus by telling jokes, playing karaoke, and answering Fresno State homecoming trivia questions.

The Hub of Digital Transformation is committed to providing students real-world experience with technological innovation, including AWS Cloud services. In 2019, they took on an innovation initiative centered on AI and robotics to improve convenience and accessibility for students. The Bulldog Autonomous Bot team, which includes current students and recent graduates, aimed to develop an application to improve awareness and usage of different campus resources and services in a new digital medium that could communicate intuitively with any user.

“I already knew how to work with AI and image processing, but never got to use it in real-world projects. With AWS RoboMaker, I was finally able to build real-life applications,” said Shifatul Islam, teammate and Lead Intern Developer at Technology Services, Fresno State. “AWS RoboMaker is the future coding environment for robotics. It makes the development process much easier, because we don’t have to worry about our laptops or device performance.”

Fresno State Bulldog Bot

The team documented their workflow to build the Bulldog Bot including:

• For hardware and physical design: The team used Husarion ROSBot 2.0, with a 3D printer for customization.
• For development: The team used AWS RoboMaker, a cloud robotics service to develop, test, and deploy intelligent robotics applications. Students completed the AWS RoboMaker Badge on AWS Educate and leveraged Robot Operating System (ROS) packages, like a remote-control node to move the robot around and a video streaming node to send a live video feed to the application. They also used software like WebSocket for fast communication and Nginx for application performance.
• For the Bulldog Bot voice: Amazon Polly turned text into life-like speech and Amazon Lex built conversational interfaces. A voice-enabled remote control allowed for voice-controlled movement.
• For navigation: The application was implemented to have remote accessible graphical user interface (GUI), providing a joystick model for the robot control and camera view.

“AWS RoboMaker has been an instrumental tool for the success of Fresno State’s new Autonomous Robots innovation initiative,” said Max Tsai, Innovation Architect and Coordinator of Digital Transformation. “This service provides a seamless platform for intelligent robot development, simulation, deployment, and fleet management with the only requirement being a web browser – including the support of the processor hungry AI integrations.”

The Hub of Digital Transformation will continue exploring opportunities for student engagement in areas of AI and ML, computer vision, and robotics to motivate and inspire students to learn and innovate in the AWS Cloud. Recognizing a learning curve that instructors and students face in learning cloud computing and developing robot applications, the Hub of Digital Transformation plans to extend these efforts to offer AWS Cloud competency workshops for the Fresno State community. These workshops will be available to students, faculty, staff, and community members across the Central Valley who have an interest in exploring AWS RoboMaker and other AWS services.

To learn more about using AWS RoboMaker in academia, check out the AWS Online Tech Talk “Robots and STEM: AWS RoboMaker and AWS Educate Unite!” Questions? Please contact aws-robomaker-academics@amazon.com.