Company knows when a machine will break before it happens


By Grove Potter republished from UBNow.

Published September 13, 2017

“I tell my friends, you can start a business in a garage in California or in a mansion in Buffalo.”
Ward Thomas, president and CEO
Sentient Science

What if someone designed a computer program that could predict, with astonishing accuracy, when a machine would fail and how it would happen? Businesses could act to make repairs before things break, saving lots of money by extending the lives of their machines.

That is precisely what Sentient Science has done, and the company’s future looks very bright.

The company, which was lured to Buffalo by UB President Satish K. Tripathi and other business leaders, is focused on the wind turbine industry, where it currently monitors about 20,000 turbines around the world. The company hopes to be monitoring 100,000 turbines by 2019.

The magic of Sentient Science’s program rests on its receiving live-streaming data from each machine and on its focus on materials science, knowing what the machines are made of down to the molecular level. Other companies tackling machine life issues rely purely on data, predicting failures based only on past experience. Sentient’s focus on materials science gives it the ability to predict how long individual components will last. And each prediction is specific to each machine because the company monitors how long each has been running and under what loads.

The company’s accuracy was proven by NASA. The space agency had 30 years of data on how its equipment had run.

“They gave us just the information from 1980, and we predicted the next 30 years of machine failures statistically perfectly,” says Ward Thomas, president and CEO of the company. “It was the first time in history something like that had happened. It was a big day for us.”

The company then took its show on the road, doing similar tests at Boeing, Sikorsky, GE, Siemens and others. “We knew one validation was not enough, so we went to the biggest companies in the world and did the same thing as at NASA,” Thomas says. “We’re a non-traditional startup. We had 13 years before we had our first commercial sale.”

Drawn to Buffalo

After NASA authenticated Sentient’s capabilities, and after the company was honored at the White House with the Tibbets Award for innovation, Sentient was in great demand. Several universities vied for the opportunity to host it. After learning about Buffalo’s emerging scientific business community and UB’s commitment to materials science, Thomas, a Toronto native, picked Buffalo as the place to grow the business.

“I believe we came at the right time when it was changing, and it is changed,” Thomas says. “Buffalo is fundamentally an incredible place. The materials science commitment that Satish Tripathi has invested in with materials scientists, which is why we’re here, it’s a mecca. There also are 14,000 processors (downtown at UB’s New York State Center of Excellence in Bioinformatics and Life Sciences) that we use every day in simulations. So the relationship with UB is one of strategic nature.”

The company is headquartered in UB’s Jacobs Executive Development Center (JEDC) in the former Butler Mansion on Delaware Avenue.

“I tell my friends, you can start a business in a garage in California or in a mansion in Buffalo,” Thomas says.

Lowering the cost of wind power

By extending the life of wind turbines, Sentient is lowering the cost of wind power.

“We want to lower the cost of energy so wind turbines can compete with coal,” Thomas explains. “That’s for the planet, for the young people and to attract the engineers.”

Sentient’s work is also in space on the Hubble telescope. NASA needed help with the lubrication of the gyros on the system that directs the telescope. The engineers at Sentient did not know specifically what they were working on because it was classified, but they came up with a gas lubricant for zero gravity that extended the number of times the telescope can be repositioned.

“That would have taken 10 years of physical testing, but they needed it fast and we were the technology the DoD (Department of Defense) had invested in,” Thomas says.

Future growth

Sentient had three people working in the JEDC when the company began operations here. It is up to 40 people now, and expects to double that soon. “We’ll tap out of this building at between 80 and 100, so we’re looking for other buildings. Someday we’ll probably build our own building as the downtown development continues,” Thomas says.

The company’s computer needs also are growing.

“We’re working with the university helping them raise funds for more computer processors. We also envision putting our own processors into the CCR (UB’s Center for Computational Research),” he says. “It can house a lot more processors, so between more state and federal investment for the university, and when we have an IPO, we’ll be investing back into the university as well.

“Our investment in tech is to make it more and more efficient. The university has helped make our code more efficient so it runs faster and doesn’t take up so much computing power,” he says.

CCR Director Thomas R. Furlani says Sentient is one of 22 companies using a super computer funded by Empire State Development to attract new businesses to the region. A proposal has been made to double the computer capacity, he adds.

The CCR also has a larger supercomputer dedicated to just academic uses.

Value proposition

The value of using Sentient is “life extension,” Thomas says.

“The ability to see into the future can make it possible to extend the life of a turbine from eight years to 30 years. That extra 22 years is my value proposition. There is nothing on a CEO’s desk for an initiative that can match that,” he says.

When the company goes public in two to three years, Thomas expects its valuation to be around $10 billion.

Sentient’s impact could reach far beyond wind turbines. The company is working on airplane engines and tackling the wheel-rail connection on freight trains. Thomas says Sentient may be able to reduce the friction by more than 10 percent.

The ability to test the lifespans of machines while they are being designed could also greatly speed product development and make the lifespan of a machine a more important consideration in the buying decision.

“Our tech levels the playing field so people for the first time are not buying only for price and delivery,” Thomas says. “They are buying for life.”