VIA co-founders Colin, Kate, Jackie, and Kristen worked together at a technology services firm Colin founded and later sold. After nearly a decade together, they teamed up with computational science expert Dr. Jeremy Taylor to start VIA in 2016.
VIA creates artificial intelligence software that identifies and prioritizes risks of equipment failure such as transmission towers, gas pipelines, and power plants. Access to sensitive power company data is often the biggest barrier to successful AI systems. With the help of Elemental Excelerator, VIA has developed Trusted Analytics Chain (TAC). TAC is a blockchain application that allows multiple companies to pool their data to create significant collective data assets, without physically transferring or aggregating data. This provides higher data security and better analytical results.
VIA has developed its software applications in collaboration with some of the world’s leading companies and government agencies in energy and security. After successful pilot projects with the Federal Energy Regulatory Commission (FERC), VIA counts Tokyo Electric Power Company Holdings Inc. (TEPCO) as an early investor and customer, and is currently working with PJM.
What makes them special?
99% of artificial intelligence analytics solutions on the market today rely solely on black box approaches to make predictions. VIA has developed a software engine that can identify cause and effect relationships, rather than just correlated events, and provide explanations for individual predictions. They are also solving the larger grid data access problem by using blockchain and distributed processing to keep data safe and encrypted.
Now let’s talk deployment.
Utilities are looking to increase the reliability of their T&D networks with limited budgets in the face of aging infrastructure, extreme weather, and strain on the grid from variable generation sources. In addition, remote equipment combined with manual processes to measure equipment status means that some companies have extremely small samples of data that are insufficient for robust analysis.