Vanguard Magazine

T TEChNOlOgy WATCh www.vanguardcanada.com JUNE/JULY 2017 39 centration indexes, and digital cameras. This data is then projected to a computer system that analyzes the information and changes the controls to ensure better products are created. It sounds like there are a lot of opportunities for this academic institute to collaborate with industry. When I think about machine learning, I almost immediately think about autono- mous vehicles. You may not have heard of this company, but Cohort Systems de- velops computing solutions that convert any vehicle into an autonomous vehicle. The solution is a mission execution system that supports deployment of hundreds of missions from which the autonomous ve- hicle will continuously review, select, and execute a mission that will achieve the most important objective for the robot, based upon the current knowledge of the robot's status and the environment it is operating in. The company is initially fo- cusing on perimeter security, where they can replace manned patrol vehicles with unmanned patrol vehicles, which is suit- able for base security, border security, air- ports, or power stations. It makes a lot of sense to implement ma- chine learning technology for something like perimeter security as it is labor inten- sive, expensive, can be dangerous, and is a role humans not do well at because it is just so monotonous. Unmanned patrolling costs less, is safer, provides a higher quality of service, and is more responsive to chang- es in the environment because the autono- mous vehicle is always alert and commu- nicates status instantly. Sometimes it just hurts to hear the robots are better than us in some areas, but we have to hear it. RANK Software is a Toronto-based company that is pioneering the use of big data analytics for cyber security solutions. One of RANK's differen- tiators is the use of deep learning. While the applica- tion of neural net- works is not new, it is innovative in the cyber security field due to the limited availabil- ity of labeled/ example datasets. RANK has developed an industry-first application of deep unsu- pervised learning to detect suspicious user strings. "Many security offerings provide the basics of anomaly detection and simple response whitelisting via Machine Learn- ing," said Mohan Rao, CTO, RANK Soft- ware. "At RANK, we are trying to solve the tough machine learning problems of true learned human-machine interaction, inferred decisions, and fuzzy intelligence. Our goal is to close the security skill gap through automated threat response." In the area of cyber security, machine learning and artificial intelligence technol- ogies can help companies stay protected against ever emerging cyber threats. This is especially true where instances such as the recent spread of "WannaCry" ransomware have shown the limitations in traditional methods and processes for detecting and responding to cyber-attacks. Through its solution called VASA (Virtual Advisor for Security Analytics), RANK monitors user and machine behaviour across multiple di- mensions in real time. This deeper visibil- ity allows VASA to detect compromised behaviour and enable rapid response, even if the ransomware is mutating and chang- ing rapidly. As the defence industry and its sup- ply chain face a global shortage in skilled security workers, they will need to trust technology to offer automated and/or prescriptive real-time response capabilities to keep up with the exponentially growing attack surface. Attack vectors are no longer external only; as seen with the Snowden document leaks, unauthorized internal ac- cess can be as costly and have longer last- ing impacts than external single instance attacks. RANK's machine learning and artificial intelligence are the foundation of its threat monitoring capability, which has resulted in a partnership with Lockheed Martin Canada to create innovative cyber security solutions for future Canadian and international defence programs. "The future of threat monitoring and in- vestigations of all types relies on our ability to successfully develop and deploy solutions that leverage data science, behavioral analyt- ics and machine learning to quickly identify and prioritize risk," said Niranjan Mayya, Founder & CEO of RANK Software. Lastly, a new company that I just met re- cently is Dat-uh, a Toronto based startup. The Founder and CEO, Humera Malik passionately explained that, "Using the power of machine learning, we have been able to tackle the complicated challenge of making data science scalable for industrial companies with our automated analytics platform." Malik said that the Dat-uh platform is helping defence manufacturers acceler- ate their big data strategies with real-time predictions and insights that move past the limitations of human-managed data mod- els and leverage automated Machine learn- ing models that are self-learning, based on AI. Personally, I see the opportunity for this application to be used right across the manufacturing sector in day-to-day op- erations to create cutting-edge plant op- erations and production processes, which will ensure Canada's place in developing the products and equipment of the future. Now that I find exciting and worth adopt- ing technology for. If you would like to connect with the companies mentioned in this article, go to theomx.com and search by company name. Nicole Verkindt is the technology editor of Vanguard magazine and founder and president of OMX. She is a board member of the Canadian Commercial Corporation and was recently appointed to the board of the Peter Munk School of Global Affairs. Jeremy James, President and Co-Founder, Cohort Systems

• Cover