When AI software startup Monolith was created five years ago, the goal of the company was to create a machine learning tool that relied more on the data and engineering expertise from the design engineer, rather than having to learn complex coding. Monolith appears to be well on the way to offer engineers a AI platform to solve a slew of engineering problems.

“We hired data scientists and looked at physics problems,” said Richard Ahlfeld, CEO of Monolith, in a recent interview with Design News. “If have an intractable physics problem, you can use machine learning to ingest machine data and solve problems you previously could not solve.”

Monolith’s AI tool can be viewed as a generative AI model which, based on the engineering data input by the user, creates a self-learning model that can understand and predict the performance of a complex system through model performance, under various operating conditions. The no-code AI model correlates inputs, which are design variables and operating conditions, for an end-to-end process.

AI Tool Helps Jota Sport Win Race

Ahlfeld noted that the AI tool can be adapted to engineering tasks in various industries, including automotive, industrial, and aerospace.  In one automotive application, Jota Sport uses self-learning models from Monolith to develop an AI application that was able to model tire degradation in a race car. The AI model enabled Jota Sport to optimize the car’s design, without having to evaluate extensive test track run data. Jota Sport finished first in the 24 Hours of Le Mans Race.

Related:AI Vision Software Lends Smart Hand to Help Farmers Count Sheep

The model Monolith Software helped to create was able to create a vehicle model based on test data that would accurately estimate the output of sensors typically used on the vehicles. The information from the model in turn can compute tire energy and provide live feedback to drivers during a race, enabling them to optimize their race patterns in real time.

Besides tire degradation, the AI tool also helped reduce wind tunnel test times and suspension modeling problems, and was able to account for different driver behaviors and tracks.

“These are open-source tools that can be modified to solve problems,” Ahlfeld said. “We are looking for problems in engineering design where people are doing something over and over again and people don’t understand the modeling.”

Ahlfeld gave several other examples of where the AI tool would be useful. Smart homes could potentially use to the tool to model the design of circuit breaker systems, including when they overheat and where to best deploy them. Cooling design in equipment and buildings is another, where the AI tool can model potential hotspots and determine the airflow required to achieve adequate cooling.

Related:AI Just Nabbed the Future of Design

Ahlfeld hopes to adopt the AI tool for emerging technologies such as 3D and the metaverse. He is also interested in adapting the platform to solve climatic and energy issues.

Spencer Chin is a Senior Editor for Design News covering the electronics beat. He has many years of experience covering developments in components, semiconductors, subsystems, power, and other facets of electronics from both a business/supply-chain and technology perspective. He can be reached at [email protected].