AI value prop

Pattern Discovery Technologies (PDT) had a problem – a big problem. It wasn’t just any big problem. It was a big data problem.

The team at the Waterloo-based data analysis company is expert in mining and analyzing big data sets. They have created software tools that allow clients to dig deep into their data and unearth previously unseen patterns and insights that can help to increase efficiencies and bolster the bottom line.

PDT’s proprietary technology has been used extensively in the oil sands industry, but now CEO Paul Sheremeto and his team are creating a new software toolkit called AssetInsight for the equipment reliability and maintenance sector. The goal is to improve reliability of equipment and to predict equipment failure, avoiding costly shutdowns.

So while PDT is able to nimbly extract, transform and load large disparate data sets, as well as “slice and dice” the data to uncover interesting insights, they are continually looking for ways in which to present the data visually in format that’s easy to digest for everyone, from a high level engineer or scientist to a maintenance operator.

Enter Communitech DATA.BASE, an initiative designed to bring about the next generation of market-driven innovation within data services. Communitech DATA.BASE is a collaboration of market participants who are exploring technology-driven ways to capture and commercialize big data in various markets. PDT was a perfect candidate to participate, given its big data dilemma.

After initial discussions, Sheremeto was eager to start building a web-based visualizer for the AssetInsight application with DATA.BASE and Communitech’s Apps Factory.

“We really saw this as an experiment,” Sheremeto says. “We believe we will find value in the process.”

Creating the first iteration of the visualizer during the schema phase of the project was highly valuable, he said, because “It really forced us to have to articulate what the process actually should be, what are we actually trying to show. It forced us to organize ourselves and actually understand what creating a web front-end to present our analysis in a simple and meaningful format had to look like and had to be.”

The wireframes have been completed and a second stage of development is under way, during which PDT will obtain trial data sets to use for building and testing the visualizer.

Sheremeto hopes that continuing to work with DATA.BASE will bring about his ultimate goal. “If the schema building process really works and lays out the structure, it will be easy to lay it out for a web developer in order to build the prototype,” he said.

Sheremeto has been presenting AssetInsight at conferences and select client engagements and is working with Invensys, a global technology company that works with a broad range of industrial and commercial customers to design and supply advanced technologies that optimize operational performance and profitability. He is eager to unveil the visualizer in the future as a value-add to clients of the AssetInsight product offering.

“And AssetInsight is just the test ground,” he says. “What we learn and develop here can be replicated to the other Insight line of tools we would like to offer. We are very excited about this. We have lots of data, we have lots of scientists, [but] we don’t have an abundance of web developers so this project will help us prepare to do some web development.”

PDT is looking forward to finalizing the schema building process in the next few months. When asked about how they will proceed to web development, Sheremeto says, “We would love to continue to prototyping with DATA.BASE and the Apps Factory. When it comes to building the real product we will definitely lean on their expertise for guidance in securing the right developers.”

While PDT are experts at discovering what they call “patterns that matter”, they look forward to being able to help their clients see those patterns as well, creating value and profit from their big data.