Amazon DeepRacer & Machine Learning at WAB

Amazon DeepRacer & Machine Learning at WAB

Alongside several initiatives promoting robotics at WAB, students are also exploring machine learning and artificial intelligence.

Amazon DeepRacer is an international program in which students learn to build and compete models that “teach” a car to drive itself around a track – similar to the technology used in autonomous driving trials. Students who are members of the WAB Amazon DeepRacer club can explore and learn on the digital platform, join an international league, submit their models into a virtual race, and even win an internship at Amazon Web Services. 

Machine learning is of growing importance in the futures of many industries. Students recently spoke with WAB alumni working in marine biology and training an underwater vehicle to explore the seabed autonomously. 

“DeepRacer provided an opportunity for me to truly delve into the possibilities of artificial intelligence and machine learning. It was a unique opportunity; no other clubs in the school present the learning possibilities presented in this club,” Grade 6 student Jayo J. said. “Since joining, I have learned much about the mechanics of the machine learning, and the many prospects it has in the real world.” 

Other applications are highly practical and incorporates several subjects, like mathematics, design, robotics, and beyond. 

Stephen Taylor, WAB's Director of Innovation for Learning and Teaching, serves as the club's supervisor.

“Introducing DeepRacer has been an interesting and practical way to engage with machine learning principles at a level that is accessible to students," Stephen said."The potential impact of machine learning and AI on the future of the workforce and industry are substantial, and this can help students make sense of and learn to master these skills. "

Machine learning is different – yet complementary – to robotics in many ways. In robotics, students design a code which dictates the robot’s precise actions. In machine learning, students develop re-enforcement learning parameters by coding in a reward system which allows the machine to learn how to complete a task with increasing success. In this case, the rewards come through successful completion of laps by the autonomous vehicle. As students train and evaluate their models, they can improve its performance.  

“I have learned that machine learning involves a lot of trial and errors, creativity, and critical thinking,” Grade 11 Ji Woo said. “I hope to continue making models that hopefully beat my own records and also challenge myself to bring good results in the monthly races.” 

Keep an eye out for more Amazon DeepRacer at WAB. Students are aiming compete in the Beijing Final, and we are looking forward to seeing the development of this, and other areas of future-focused technology, in the coming years. 

  • STEM