NON-EXPERT PRACTICAL APPLICATION OF AI VISION SYSTEMS IN DESIGN ENGINEERING PROJECTS
Authors: Garland, N., Wade, R. and Palmer, S.
Journal: Proceedings of the 25th International Conference on Engineering and Product Design Education: Responsible Innovation for Global Co-Habitation, E and PDE 2023
Pages: 7-12
Abstract:Design projects units for BSc (Hons) Design Engineering students at Bournemouth University integrate and apply knowledge from a range of taught units together with self-directed learning and towards solving design problems. Recently, level 6 (FHEQ) project students have proposed and designed solutions that require AI vision-systems. These projects presented a problem for supervision, with limited, or no expertise in the technology or available equipment; students therefore treated these subsystems as a “black-box” exercise. To address these issues a set of technical requirements were compiled, a range of AI technology solutions were identified before selecting the Nvidia Jetson Nano. From the literature, a stream-lined practical program was developed to introduce the technology to level 5 and level 6 project students as part of their design education. This provided hands on experience through familiarization with the interface and the use of pretrained models before students re-trained networks with their own datasets. Level 5 students utilised the technology to develop a scratch detection machine for sorting damaged components. Level 6 students were provided with the opportunity to integrate the technology into projects where appropriate and two students did so; one developed a device to identify people trapped in buildings after an earthquake, the second developed a device for monitoring chili-plants when grown under polytunnels. Developing and delivering the introductory programme as a non-expert learning pathway has enhanced the student experience within design education, provided a simple workflow that students can utilise and build upon, and led to successful student outcomes.
https://eprints.bournemouth.ac.uk/39679/
Source: Scopus
Non-expert practice application of an AI vision systems in design engineering projects
Authors: Garland, N., Wade, R. and Palmer, S.
Conference: E&PDE 2023: 25th International Conference on Engineering & Product Design Education
Pages: 1-6
ISBN: 9781912254194
Abstract:Design projects units for BSc (Hons) Design Engineering students at Bournemouth University integrate and apply knowledge from a range of taught units together with self-directed learning and towards solving design problems. Recently, level 6 (FHEQ) project students have proposed and designed solutions that require AI vision-systems. These projects presented a problem for supervision, with limited, or no expertise in the technology or available equipment; students therefore treated these subsystems as a “black-box” exercise. To address these issues a set of technical requirements were compiled, a range of AI technology solutions were identified before selecting the Nvidia Jetson Nano. From the literature, a stream-lined practical program was developed to introduce the technology to level 5 and level 6 project students as part of their design education. This provided hands on experience through familiarization with the interface and the use of pretrained models before students re-trained networks with their own datasets. Level 5 students utilised the technology to develop a scratch detection machine for sorting damaged components. Level 6 students were provided with the opportunity to integrate the technology into projects where appropriate and two students did so; one developed a device to identify people trapped in buildings after an earthquake, the second developed a device for monitoring chili-plants when grown under polytunnels. Developing and delivering the introductory programme as a non-expert learning pathway has enhanced the student experience within design education, provided a simple workflow that students can utilise and build upon, and led to successful student outcomes.
https://eprints.bournemouth.ac.uk/39679/
Source: BURO EPrints