Get Appointment

  • contact@wellinor.com
  • +(123)-456-7890

Discovering new Shape Memory Alloys (SMAs) using AI

shape memory alloys

Applying the power of data intelligence to discover smart materials and innovate new stimuli in shape memory alloys with targeted properties

Problem Statement:

With the increasing usage of SMAs in various industries, there is always a trend to find new multicomponent NiTi-based shape memory alloys (SMAs). But Ni-Ti alloys are expensive. The problem is identifying Ni-Ti alloys with the targeted property of low transformation temperature (TP) and as economical as Cu and Fe Alloys.

shape memory alloys

Challenge

To build, train and deploy machine learning models in combination with global optimization techniques to effectively navigate the complex search space and find new materials with targeted properties.

Solution

We employed an adaptive design loop that uses a trade-off between exploration and exploitation of the results from our regression model to guide the experiment that needs to be performed to explore the total experimental area and focus on a local area with the apparent global optimum.

The steps involved in the loop are –

  1. An initial experimental data set of 53 NiTi-based SMAs with known transformation temperature, and features serve as input to the inference model.
  2. The model is trained and cross-validated with the initial alloy data.
  3. The trained model is then applied to a data set of unexplored alloys, i.e. the virtual dataset.
  4. The design automatically chooses the best candidate for synthesis and characterization.
  5. The measured transformation temperature augments the initial data set to improve the inference and design further.
shape memory alloys

We employed different design functions or selectors based on a heuristic referred to as global optimization, which has been extensively used in the aircraft and automobile industries as surrogate-based optimization. For new materials discovery, this optimization allows us to choose potential candidates for experiments based on maximizing the ‘expected improvement’ over the search space.

Impact

The adaptive design framework we applied helped us in discovering new alloys with targeted transformation temperatures. The same strategy can be applied to find out new materials with other desired properties.

Product: Material discovery using AI

Company: India-based material research organization

Location: Hyderabad, India

Recent Case Studies

digital menu app

We harnessed the power of AI to build a digital-age solution for a restaurant business in allowing their consumers to browse digital…

ecommerce application

Our expertise leveraged their skill in building a next-gen eCommerce application for our client to serve the future need of business…

marketing web application

Discover how we worked towards achieving an easy-to-use, intuitive, and inviting UX/UI for our client’s website to sustain digital age challenges.

Ready to get Started?

Let’s connect one-to-one for insight-driven business results.

Product: Material discovery using AI

Company: India-based material research organization

Location: Hyderabad, India

    Get In Touch

    [mathcaptcha mathcaptcha-401 class="form-control"]

    By Submitting this form, I agree to "Soulpage" Privacy Policy & Terms and Conditions

      Talk to us


      By Submitting this form, I agree to "Soulpage" Privacy Policy & Terms and Conditions

        Book a Demo

        By Submitting this form, I agree to "Soulpage" Privacy Policy & Terms and Conditions

          Download Usecase For Free