TORONTO – Telus announced today it has partnered with the Vector Institute to launch the Energy Optimization System (EOS), which is designed to help reduce operational costs and electricity consumption in data centres across Canada using AI technology.
“This collaborative development uses model-based reinforcement learning (MBRL) to fine tune the heating, ventilation, and air conditioning (HVAC) systems across network locations, allowing for energy-efficient temperature control,” a press release explains.
“Annually, an estimated 40 per cent of energy consumed across Telus network locations is directed towards cooling telecommunications equipment, which is critical to maintaining optimal network performance,” according to the release.
“Telus team members collaborated with the Vector Institute to build and test a solution to reduce this number and create a more sustainable use of HVAC equipment in 24/7 data centres.”
A pilot test showed a nearly 12% decrease in annual electricity consumption in a small data centre.
“We wanted to find the most efficient opportunity for temperature control in data centres, one that considered both the environmental and economic costs, while also providing the best service for our customers,” said Jaime Tatis, vice-president, data strategy and enablement at Telus, in the release.
“Using MBRL to develop an AI agent that chose specific actions based on rewards, we were able to allow for quick learning by machines to create new effective and efficient policies for controlling HVAC systems.”
Telus and Vector decided to open source the algorithm used “as a contribution to the energy conservation community, furthering commitments by both tech leaders to leverage AI to create better outcomes for Canadians and our environment,” the press release says.
Telus is a founding sponsor of the Vector Institute and is “committed to using this partnership to advance the AI ecosystem across Canada, producing an economic boost and creating meaningful impacts with technology.”
For more, please click here.
