AI and Machine Learning

In a world of increasingly complex challenges, our experts are using artificial  technologies and machine learning as integral tools in nearly every area of mechanical engineering.

Machine learning and AI

  • People running a race around a track

    Safe AI

    The idea of self-driving cars or robot aids to help around the house was once a dream of the future. Today, these things are a reality and our faculty are using machine learning to make human-robot interaction safer, and smarter.

    Faculty involved: Jonathan Cagan, Amir Barati Farimani, Eni Halilaj, Philip LeDuc, Conrad Tucker, Ding Zhao

  • People running a race around a track

    Human health

    Machine learning can help us to improve human health in many ways, like predicting and preventing musculoskeletal injuries, personalizing rehabilitation, and developing antibodies to thwart quickly-mutating pathogens.

    It can also help us to enhance the analysis of medical imaging, model the complex geometry of neurons, design synthetic biological systems, and provide a patient's vital signs to a medical team before arriving at the hospital. Our mechanical engineers bring their unique expertise to these biomedical challenges.

    Faculty involved: Jonathan Cagan, Amir Barati Farimani, Eni Halilaj, Philip LeDuc, Conrad Tucker, Jessica Zhang, Inseung Kang

  • glass flash with pipettes

    Material discovery for energy applications

    Safer, more robust batteries... new ceramic polymer hybrid materials that require less energy to produce... heat transport prediction and improved energy conversion... These are some of the areas our faculty are investigating with the assistance of machine learning tools.

    The computational power of machine learning can screen and reject millions of possible combinations that allow researchers to hone in on the best solutions. This process would otherwise take decades of trial and error in a laboratory setting.

    Faculty involved: Amir Barati Farimani, B. Reeja Jayan, Shawn Litster

  • 3d printed parts

    Design and manufacturing

    With artificial intelligence and machine learning, our experts are transforming and optimizing design and manufacturing. Here are a few examples: creating new concepts for cars and aircraft with design DNA; using computer vision to detect flaws during 3D printing; turning static drawings into active simulations with smart design tools; and developing virtual reality engineering simulations to place students in an interactive, immersive manufacturing environment.

    Faculty involved: Amir Barati Farimani, Jack Beuth, Jonathan Cagan, Burak Kara, Christopher McComb, Sneha Prabha Narra, Kenji Shimada, Conrad Tucker, Jessica Zhang, Inseung Kang

  • aerial drone in a canal

    Infrastructure, smart cities, and society

    Inspecting canals and power plants with drones, mapping air pollution and calibrating low-cost air quality sensors, desalinating water with new, energy-efficient materials, and using social media data to predict threats... these are a few examples of how machine learning can play a role in maintaining and improving the infrastructure in our communities.

    Faculty involved: Amir Barati Farimani, Albert Presto, Kenji Shimada, Conrad Tucker

    • People

    Learn who in Mechanical Engineering is working in maching learning and artifical intelligence.

    Related news

    Mission accomplished Opens in new window

    CMU engineers and scientists undertook more than 45 research projects to develop artificial intelligence approaches to enable the use of metal additive manufacturing for the U.S. Army.

    Diversifying DNA origami Opens in new window

    A new computer aided design tool from the labs of Jon Cagan and Rebecca Taylor enables scientists to rapidly generate and manufacture novel DNA origami structures within designer-defined constraints.

    Making way for a digital twin of the human placenta Opens in new window

    A computational model of the basic functional unit of the placenta enables researchers to simulate placenta blood flow, realistically.

    Transformer-based models for identifying alloy properties Opens in new window

    Researchers develop AlloyBert, a modeling tool designed to predict the properties of alloys.

    The trend is your friend … and so is machine learning Opens in new window

    Grasping both general market trends and finer details, a new, hybrid machine learning model from researchers in the Department of Mechanical Engineering predicts financial market volatility with increased accuracy.

    Multimodal machine learning model increases accuracy Opens in new window

    Researchers have developed a novel ML model combining graph neural networks with transformer-based language models to predict adsorption energy of catalyst systems.

    Putting the ghost in the machine Opens in new window

    To build AI teammates that better reflect human designers, Chris McComb is emulating cognitive styles in large language models.

    Safe AI for safe streets for all Opens in new window

    Ding Zhao is teaming up with the City of Pittsburgh to improve road safety and eliminate traffic fatalities using AI.

    Explore Other Research Topics