Artificial intelligence (AI) is causing a great deal of excitement among researchers and scientists, not to mention large corporations. On the other hand, many small and medium enterprises are still hesitant to jump on the bandwagon. Are you of the opinion that artificial intelligence is only accessible to large companies? Think again! Québec SMEs like Distech Controls and DIAGNOS have taken advantage of the latest advances in AI by calling on the expertise of researchers from École de technologie supérieure (ÉTS).
Professor Marco Pedersoli is looking to reduce the cost of computing and the quantity of data required for machine learning, which is exactly what was needed at Distech Controls, a company that offers innovative solutions for connected buildings. According to Professor Pedersoli: “One of the main problems with deep learning is that it requires massive computing resources and data to accomplish the learning.” Given that most of the engineers working at Distech are ÉTS graduates, it was only natural to call upon the expertise its researchers. That is how Steve Lupien, vice president, Building Automation Technology at Distech, and Marco Pedersoli and Éric Granger, both professors at ÉTS and members of the Imaging, Vision and Artificial Intelligence Laboratory (LIVIA), laid the groundwork for a joint project that is expected to continue over the next five years.
Working in collaboration with the university milieu gives companies access to the latest knowledge related to AI, because it is not always possible for them to accomplish their goals internally.
Recognizing the immense potential of AI for his sector of activity, Steve Lupien began exploring the possibilities with the company’s employees. “After a brief feeling-out period, it quickly became clear that we would be better off collaborating with experts like Marco and Éric,” he explains.
Moreover, this type of collaboration gives companies access to a highly specialized labour force that already understands their needs. According to Éric Granger: “This project allowed us to obtain funding to hire students. We provide them with training and they work on real projects, and eventually become potential employees for our company.”
To assist Distech in integrating AI into its solutions, the research will focus on two themes: optimizing spaces with a view to facilitating the cleaning, ventilation or use of offices, for example; and improving eco-friendliness with a view to reducing energy consumption as much as possible. Mr. Lupien specifies: “Artificial intelligence will enable us to automate many of the tasks that were previously performed manually. It will allow us to simplify our products and enhance the comfort of occupants.”
In the case of DIAGNOS, the company employs a dozen or so individuals and has been working for more than two years on a joint project with Professor Ismail Ben Ayed, who is also a member of LIVIA. Although this SME has been using AI for the past 15 years, it doesn’t have the expertise required to take advantage of deep learning. “That’s where Ismail comes in,” explains André Larente, president of DIAGNOS. The reputation that ÉTS and LIVIA enjoy enables them to attract extremely talented researchers, including Adrian Galdran, a post-graduate student who has been working on the medical imaging project with DIAGNOS for a year now.
The results are tangible. In a matter of seconds, rather than minutes, the application can now analyze a photograph of the retina in order to determine whether the patient’s vision is at risk from a disease such as diabetic retinopathy.
According to Mr. Larente, who would like to continue the collaboration with other applications: “Our goal is to help physicians by giving them more effective tools.”
Professor Ben Ayed affirms that “this type of collaboration is extremely important. My main goal is to see my knowledge transferred and impact people’s lives. Otherwise, it’s just a bunch of equations on paper.
