Launched out of the University of Cincinnati, Predictronics Corp. delivers artificial intelligence (AI) based predictive analytics for industrial applications. The company’s data reduces downtime, optimizes productivity and improves product quality. Cincinnati Future spoke with Predictronics’ three founding members—Patrick Brown (chief financial officer), Edzel Lapira, PhD (chief executive officer) and David Siegel, PhD (chief technology officer)—about what the company does.
Tell us about the company.
Brown: Our company is a spinoff from UC. We have a research lab here at UC that focuses on AI, machine learning and focused solutions for predictive maintenance in manufacturing. We are creating tools for predicting when a machine will fail so that companies can prevent that failure and reduce their unplanned downtime.
It’s important to know what has gone wrong when failures occur and to know when to bring in maintenance at an opportune time so you can still meet production requirements. Usually that is done with a manual inspection. We’re transforming the traditional practices for maintenance and quality assessment to use a more predictive, data-driven approach.
Siegel: The general idea is to provide a solution that reduces unplanned downtime by finding early indications of those failures. Catching just one of those failures early can save a company hundreds of thousands of dollars or maybe more, depending on their business case.
How did the company get started?
Brown: Edzel, David and I spun the company off, along with our lab director Professor Jay Lee. We started as a consulting company, but we realized that scaling this kind of solution is difficult to do on a consulting basis. It’s a very specialized skill set—mechanical engineering, manufacturing and industrial knowledge, combined with data science, machine learning and AI algorithms. Finding people with those skills and then also training them is difficult.
So, we decided to create our own platform that speeds up the process of developing solutions for our customers. We deploy those solutions and then sell software subscriptions to our platform. That’s our business model.
How does the platform work?
Brown: We collect data, such as vibration, temperature and speed, from the customer’s operation. We then analyze the data, build the underlying machine learning and AI health model for whatever asset we are realizing. We build a model and deploy it using our end-to-end software platform. That covers every part of the solution, from data ingestion to analysis to a visualization component that we monitor. Then, we send reports to our clients to let them know when a failure might be coming or when a bad part is being produced.
Give us an example of your product in use?
Brown: A good representative case would be industrial robots. Let’s say you have a welding robot that is responsible for constructing a car door by putting two pieces of metal together. If that robot were to fail, then it would stop the entire production line. With most automotive companies, every car that comes off the line is already sold to a dealer. So any time they can’t fill the order on time, they have lost revenue. Not to mention the time lost by people on the shop floor standing around waiting for the process to come back up.
Your company came out of UC. Can you talk about some of the benefits of being located in Cincinnati?
Brown: One benefit of having a company in Cincinnati is that we have a constant supply of great software developers, mechanical engineers and people with the experience that we’re looking for. We can still acquire talent from the research lab we split off as those students graduate. Also, we’re in the Midwest, which is known for manufacturing. So we have people who have that kind of manufacturing mindset.
Lapira: People here understand manufacturing. They understand the value of manufacturing, and they understand those principles. Another benefit is that we are a central hub for our customers, who come from Michigan, Kentucky and Tennessee. We are smack dab in the middle.