Winners of the 2020-2021 William and Rhea Seath Awards Competition


Professor Nathalie Tufenkji and Dr. Mathieu Lapointe, postdoctoral fellow, Chemical Engineering

Executive Summary

With an annual global market of $18 billion, coagulants and flocculants are critical to water treatment but carry an economic and environmental burden. For wastewater alone, these chemicals generate ~8 million tons of metal-containing sludge waste annually. To simultaneously deal with the issues of process sustainability, cost, and efficiency, we have developed new materials notably reengineered using recovered waste from treatment plants; namely, cellulose, polyester, cotton, and keratin fibers. These advanced materials (functionalized fibers and microspheres) drastically improve removal of conventional and emerging contaminants during settling by increasing floc size and density. Moreover, we developed a three-in-one bridging/ballasting/adsorbing cellulose-based material (flake) that can simultaneously adsorb contaminants, bridge colloids, and ballast flocs, whilst reducing chemical usage. The unprecedented size of flocs produced using flakes enables easy floc removal by screening, eliminating the need for a settling tank, a large and costly process unit. These reusable materials combined with separation via screening will allow global water treatment facilities to reduce their capital and operating costs as well as their environmental footprint. Because of their relatively low cost and unprecedented performance, we expect that our products will be commercialized globally. In fact, our materials would be notably used during aggregation-settling, a technology used by more than 70% of the water treatments plants in the world. This award will be used for salary support and the fabrication of a portable pilot unit to test the materials for different market applications as we drive the technology to commercialization. This award will significantly accelerate our market entry by enabling us to demonstrate the performance of our disruptive technology at a larger scale and over a wider range of operating conditions (i.e., for several industrial sectors requiring water treatment).

Profile Photo of PavelProfile Photo of Ioannis PsaromiligkosProfile Photo of Zeljko ZilicPOWER-EFFICIENT AI-PROCESSOR WITH DIRECT RAW CAMERA-SENSOR DATA PROCESSING

Pavel Sinha, PhD Student, Professors Ioannis Psaromiligkos and Zeljko Zilic, all Electrical and Computer Engineering

Executive Summary

Artificial Intelligence (AI) is an emerging technology that is slowly making its way into our day-to-day life. Currently, AI is mostly in the cloud computing space, but in the future, AI will be present in every electronic device. Like the microprocessors penetrating every electronic component since the early 1990s, in the future, dedicated AI computation circuitry will be into every electronic device regardless of shape and size. Companies trying to build such computing devices are presently concentrating on the hardware component, which has been the traditional approach. We believe that the optimal approach is to simultaneously optimize the algorithm, hardware platform, and the overall system, resulting in a scalable and highly cost-optimized solution. We have a patent-pending AI-Processor hardware architecture that is best in its class for power consumption. The technology enables running on battery power for over a year without needing a battery replacement. We have a patent-pending technology to integrate our AI-Processor in most existing systems without the need for hardware re-spin, unlike our competition.

Further, we differentiate ourselves with patent-pending technology that enables the application of optimized AI algorithms directly on raw camera sensor data, reducing the need for expensive image signal processing hardware, thus lowering cost and power consumption. Finally, our integrated software development platform enables developers to effortlessly integrate AI algorithms from any of the popular Python/Matlab environments to our AI-Processor. We have demonstrated several attractive applications such as real-time object detection and classification and automatic lip-to-audio synthesis. We are currently in talks with potential customers that have expressed sincere interest in our technology and are in an engineering collaboration with one of them. We are currently developing the integrated circuit (IC) in an industry-academic partnership with McGill University and CMC. We intend to use the WRSA funds to pay the salary and stipend for interns and pay for material costs for the AI-Processor development and testing.

Back to top