By Beth Gallagher
Communications & Public Affairs
![Professor Keshav](/computer-science/sites/default/files/uploads/images/srinivasan-keshav-380x320-267x225.jpg)
But Srinivasan Keshav, a computer science professor in the David R. Cheriton School of Computer Science saw that solving the problem could reduce greenhouse gas emissions that harm the environment.
Keshav and math graduate student Peter Xiang Gao have developed SPOT, a Smart Personalized Office Thermal control system that uses sensors to predict how comfortable you are and then automatically adjusts the heat or the fan in your own workspace.
Finding your comfort zone
“We guarantee that you’ll be comfortable all the time, says Keshav. “Your comfort level can actually be mathematically determined.”
The larger impact comes in because the overall temperature of office buildings can be set lower in the winter and higher than normal in the summer. A small change in the set point of the entire building can mean a 10 per cent savings in energy costs and a drop in energy use, says Keshav.
The impact to the environment is substantial because 40-50 per cent of energy used in a typical office building goes to heating and air-conditioning.
Using math to solve real-world problems
SPOT is just one of the projects Keshav and a team of researchers are working on at the University of Waterloo’s Information Systems and Science for Energy (ISS4E) laboratory. Researchers in the lab are using mathematical models to solve practical problems that will ultimately reduce society’s carbon footprint.
One of the projects involves a mathematical framework that enables First Nations communities to use more battery power and less diesel fuel. Remote communities are not connected to the grid and therefore rely heavily on diesel generators for power. Diesel emissions are known to be a significant contributor to global climate change.
“Even a small battery gives you a big win,” says Keshav. “It still reduces emissions by allowing diesel generators to run at maximum efficiency.”
ISSE researchers have also found a way for solar farm owners to accurately predict how much energy they will be able to produce so that they can bid on and win contracts to supply energy to the electrical grid. Historically, renewable energy sources have found it difficult to compete with nuclear, hydro and coal plants that can easily manage and predict how much power they can produce.
“Now solar farms can be very sure – 999 times out of 1,000 – that they will be able to meet a supply contract. They can be a player.”
Projects at the ISS4E laboratory operate under a guiding principle: “We are trying to reduce the carbon footprint of energy systems,” says Keshav.