About the role#
This internship focuses on building machine-learning tools to support perovskite process development and quality control. You will work alongside an engineer mentor, using internal imaging tools and data libraries to improve defect detection and recipe optimization. The project spans two phases over six months. In the first phase, you will train a transfer-learning computer vision model to identify defects in perovskite device images. In the second phase, you will build a Gaussian Process surrogate model using Ax/BoTorch to propose experimental parameters. You will also have the opportunity to extend the defect library and implement a statistical process control framework. At the end of the term, you will present your project summary and accomplishments to the leadership team.
What you'll do#
- Train a transfer-learning computer vision model to identify device defects.
- Build a Gaussian Process surrogate model to suggest next experiments for coating and annealing parameters.
- Develop a closed-loop optimization framework to replace traditional search methods.
- Extend the existing defect library.
- Apply a statistical process control framework to classifier outputs.
- Present your work to leadership.
What you'll need#
- Current enrollment as a junior, senior, or graduate student in Computer Science, Data Science, Statistics, Electrical Engineering, or Computer Engineering.
- Proficiency in Python and standard machine learning frameworks like PyTorch, TensorFlow, or scikit-learn.
- Experience with image data, including CNNs and transfer learning.
- Familiarity with Bayesian optimization or Gaussian Process modeling, such as Ax/BoTorch.
- Experience with OpenCV or similar image-processing libraries.
- Understanding of basic statistics and experimental design.
- Ability to build and document reproducible data pipelines.
- Strong communication skills.
- Background in materials science, thin-film, or photovoltaics is a plus.
Location & details#
- Location: Perrysburg, Ohio.
- Term: Winter 2027.
- Modality: On-site.
- Employment: Full-time paid internship.
About First Solar
First Solar operates as a public company specializing in thin-film solar panel manufacturing and renewable energy. Founded in 1999, the firm maintains its headquarters in Tempe, Arizona. It employs approximately 6,715 people across its global operations. The company develops its solar technology through research and development labs located in California and Ohio.
How to get in at First Solar
Applying to First Solar early is a distinct advantage because recruiters review applications as they arrive. Intern Insider sends an instant alert the moment a role matching your target is published, helping you get your materials in before the pile grows. You gain a better chance of being seen when you are among the first to apply. Reaching the recruiter directly often beats applying into a standard queue. Intern Insider surfaces the specific recruiters behind roles at First Solar so you can reach out to ask about the position or a referral. This direct approach materially improves your response rates compared to a blind submission.
