AI Self-Driving Golf Cart Now Ready to Offer University Visitors Wonderful Tour
Release Date 11 July 2022 Mandarin
NKUST has formed an AI Golf Cart project team with Microsoft, Turing Drive, and ADLINK Technology, Inc. (ADLINK) to run the University’s artificial intelligence (AI) systems on the Microsoft Azure cloud platform and complete the first Level 2 autonomous driving golf cart (Golf Cart). The Golf Cart is designed to operate on the campus and could offer University visitors a wonderful campus tour in the future.
The Microsoft Azure cloud platform enables students and teachers across different departments, schools, and colleges to develop, compute, and optimize their models or systems on the same platform. In this case, the platform allows up to 30 people to tune the same model at the same time, and this feature expedites the project's progress. The platform as a service (PaaS) environment in the cloud enables the University to boost the training efficiency of the Golf Cart, which is also the first NKUST autonomous golf cart that uses Azure AI.
In addition, this AI Golf Cart project exemplifies the importance of making AI courses become general education for the University.
“NKUST has been investing in the fields of the AIoT along with the ICT for a long while...self-driving or autonomous driving car technology...will be our future focus to cultivate students with skills of application development and relevant knowledge.”
“NKUST has been investing in the fields of the AIoT along with the ICT for a long while. We are glad we have many achievements and successful efforts that applied 5G and AIoT in smart manufacturing. Also, we will not miss out the future trend, that is, self-driving or autonomous driving car technology. That will be our future focus to cultivate students with skills of application development and relevant knowledge,” says President Yang Ching-Yu.
He adds that the University has been dedicated to making computer programming education available for all students as well as building many innovative teaching practices and facilities. These courses help students to build and run their programs on an AI robot and adjust their design based on real-time feedback from the robot.
This AI Golf Cart project runs in the PaaS environment on the Microsoft Azure cloud platform, which enables students of different majors to work hand in hand with project teams.
This project is a game changer for the University because it shows how technologies, knowledge, and skills can be shared, collaborated, and constructed on a cloud platform across a variety of fields and industries in a project.
Dean of R&D Dr. Kuo, Chun-Hsien (who currently served as Vice President since Aug 1, 2022) stresses the point that the educational goal for the University is to make AI courses become general education.
Besides the use of the Microsoft Azure cloud platform, the University has teamed up with Turing Drive and ADLINK in order to make the operation easy to use for all learners.
The University has trained its first batch of students who are well versed in tools and services provided by Azure Machine Learning (Azure ML) and trained to use artificial neural networks model, including the use of autonomous labeling tool.
Moreover, students learned to use Docker, a set of PaaS software, simultaneously to build, train, test, and complete artificial neural networks, which are required models for autonomous driving.
The team put Autoware, an ADLINK software for autonomous driving golf carts, in Azure Virtual Machine (VM) in order to test and simulate their design in the Azure environment. By doing so, they can observe the driving direction of the Golf Cart and its trajectory.
The first autonomous driving model developed by NKUST was designed for outdoor navigation. This model enables the Golf Cart to give guests or visitors a safe and enjoyable campus tour.
In addition to this function, the Golf Cart can be a great help for the University to transport heavy goods on campus and make the campus a sustainable University.
The Microsoft Azure cloud platform provides users to calculate and analyze the geospatial movement trajectories of the autonomous driving car. It allows teams to work on the cloud at the same time and integrate numerous applications so as to cut developers’ coding time.
These features make talents among a variety of departments or colleges exchange their experiences and develop this successful case, being a model for future projects.
Translated/Edited by Jess Lin