“You can see it…like a blanket over the city.”
Aiden Livingston was inspired to start ThermoAI, based in St. Louis, as a response to the pollution he encountered while running tests in South America for a company he founded in New York. Livingston recounted that the industrial emissions he experienced in Bogota, Colombia and in Santiago, Chile were thick and oppressive.
“I had never been exposed to anything like it in the U.S., and I began to develop asthmatic symptoms from all the pollution. There’s so much heavy industry in South America, and almost no regulation around clean air,” said Livingston. “Air pollution associated with heavy industry is very bad. The manufacturing processes are inefficient, and the result is detrimental not only to air quality and health, but also to the bottom line for those operations. It’s not only an environmental and health issue…these inefficiencies obviously impede the ability for production facilities to achieve optimal results.”
Livingston teamed up with data scientist and machine learning specialist David Watson to start ThermoAI. They recruited Carolina Chaves, a sustainable energy solutions researcher and chemical engineer as ThermoAI’s COO. Chaves’ work is focused on optimizing combustion-based products. David Watson is CTO at ThermoAI, and a doctoral candidate at the University of Oxford. Livingston has authored the books The AI Entrepreneur: How to bring AI to Legacy Industries and Saving the World on a Shoestring. ThermoAI was a 2018 Arch Grant recipient.
ThermoAI has developed AI technology and Machine Learning to optimize operations in the energy industry. The company works with large energy companies in the US, South America, and Canada. The ThermoAI platform analyzes data and various interactions to determine the optimal conditions needed to achieve near-perfect combustion.
“Combustion requires an exact mixture of air and fuel. Power plants are using just a standard table to determine the ratio of fuel to air to run large boiler systems. However, this ratio is dependent on many variables such as weather and humidity that change minute-to-minute,” said Livingston. “This mode of operation is inefficient because it relies heavily on human judgment to determine whether the mixture is too rich or too lean.”
ThermoAI uses sensors to collect data that is used to predict air and fuel mixtures. The company’s cloud-based sensors collect data that can be viewed in real time via a mobile interface.
“Our AI allows us to change levels ahead of time to increase efficiency of overall operations. We can tell plants how much fuel they are putting to productive use versus what they are losing in pollution. We are increasing efficiency from roughly 33% to 45-50%, and that equates to significant savings, and well as less pollution,” said Livingston.
The ThermoAI technology recreates the elements and operations of a plant – physical assets, processes, and systems – resulting in a virtual environment, or “digital twin” to run simulations. ThermoAI uses millions of possible configurations based on data from the sensors to find the optimal settings for any set of conditions. The platform’s dashboard provides the indicators operators can use to optimize asset performance and utilization.
ThermoAI team is focused on attracting more clients in the Midwest and is working toward growth by demonstrating other use cases beyond optimization of the energy industry. Livingston notes that ThermoAI technology has application in anything using combustion, such as paper making and metal manufacturing. The company is looking to run additional industry pilots to collect even more data to demonstrate the benefit of the company’s technology.
Visit www.thermoai.com to learn more about this company.