Walsh G37 is the site of a pilot project using AI to do audio-visual checks in teaching spaces.
In a pioneering and award-winning move, Otago is harnessing the power of AI to speed up daily audio-visual checks of its teaching and learning spaces.
Each day, audio-visual and IT staff at universities across the globe race through teaching spaces doing quick checks to ensure systems will work. The checks are necessary but time consuming.
Audio-Visual Support Team Leader Ryan Walker says that at Otago these checks have historically taken multiple staff about 25 hours a day combined.
Ryan Walker
“When I started thinking about how daily checks could be more efficient, scale up without increasing staff, remove capacity for human error and integrate with other systems, AI seemed the logical answer.”
Ryan started a pilot project, using AI to do the checks in one room, Walsh G37, which then alerted staff to any issues.
The pilot has been so successful that the technology will be rolled out to 50 more teaching and learning spaces next year, and potentially all 150 rooms within the next five years.
The pilot project won Ryan an Excellence in Technology Innovation award at the Tertiary Information Computer Technology (ICT) Conference 2024 in Christchurch, and he believes Otago is one of the first universities in the world using AI for this purpose.
Ryan has shared some AI room checking code online to help promote innovation and collaboration, and other universities have approached him to find out how exactly he is doing it.
In the pilot room, Walsh G37, the process kicks off very early in the morning.
The University’s Resource Booker system is checked to ensure no one has booked the room. An automated process uses motion detectors to check for movement to ensure the room is actually empty before AI can start its checks.
AI then turns on the room’s computer and large on-wall screens, takes a photo with a moving camera – that usually focuses on lecturers or whiteboards – so AI can analyse the photo to check if the wall-mounted screens are showing the information from the computer.
AI also checks whether speakers, microphones, and lights are working. It checks for hazards, including blocked fire exits and cables that could trip people.
If there are any issues, AI reports them to staff in Teams. This saves them from manually searching through the audio-visual systems to pinpoint the exact problem because the AI has already suggested the potential causes.
The daily checks are vital because faults can develop for no apparent reason, Ryan says.
“Even the camera taking photos for AI suddenly started showing upside down images in green after working faultlessly for a month.”
AI has also been used to confirm that various scheduled tasks are happening as they should, such as checking if the pilot room’s computer restarts at 5 a.m. as scheduled. The camera takes a photo every 30 seconds around that time, and AI reports in writing to staff what it is ‘seeing’.
“It’s amazing to get a report at 5am with all the details,” Ryan says. “I would not get a person to do this.”
His overall aim with AI is to enhance safety, efficiency, and reliability, to enrich the teaching and learning environment along with confidence in educational spaces.
“At the moment, it (AI) is doing really simple checks, with a sprinkling of intelligence, but the world’s our oyster.”
Ultimately, Ryan wants the daily checks to link into the system that manages requests for service and interactions at the University, including the AskOtago Portal.
His work fits with the University’s Pae Tata Strategic Plan to 2030 which envisages transforming the digital environment to create value for students and staff. The plan also envisages the University’s digital spaces operating seamlessly in tandem with physical spaces to enhance our staff, students’ and communities’ experience of the University, while also reducing staff’s workloads.
- Kōrero by Gail Goodger