Researchers at the Indian Institute of Technology-Madras have developed a software which they say can help solve engineering problems in fields from thermal management and semiconductors to automobile, aerospace and electronic cooling applications much faster than existing methods.
For a few years now Artificial Intelligence, machine learning and deep learning are being used in areas such as signal processing, speech recognition, image reconstruction and prediction, but few attempts have been made to use them to solve engineering problems.
Assistant Professor at IIT Madras and leader of the research team, Dr. Vishal Nandigana, said that the software has been found to be nearly million-fold faster compared to existing solutions for thermal management problems. “Our software works on any generalized rectilinear and curvilinear input geometry. Our research saves computational time, which is the bottleneck to solve most engineering problems,” he explained.
“We utilised data-driven AI and deep learning to arrive at solutions for engineering problems after training AI with data sets. These prior data sets can be from existing big data in the relevant engineering industry, where a lot of experimental data are available. Also, if data is not available for training the AI, it can be generated using commercially-available CFD (Computational Fluid Dynamics) software on small independent pieces of the full-blown problem,” he added. The team, he explained, used a novel Recurrent Neural Network (RNN) and a Deep Neural Network (DNN).
Researchers are working on establishing a startup named Alsoft to offer solutions for various engineering problems based on the new software, according to an institute press release.