Why should I participate?

Whether you are coming from microscopy or ML/AI side, hackathon is an opportunity to make real impact.

You will work on real microscopy and spectroscopy problems, gain hands-on experience with digital twin AFM/STEM simulators, and learn practical ML workflows used in research labs and industry.

You will get visibility with judges and sponsors from universities, national labs, and companies — great for jobs, internships, collaborations, and graduate school.

You also build a strong GitHub portfolio, connect with a global community across 20+ sites and online, and compete for multiple awards.

Why organize a local site?

Local sites create a stronger, more social hackathon experience. They allow participants to collaborate in person, form teams more easily, and get real-time guidance from local mentors.

For organizers, it gives increased visibility for the institution, community engagement, easier outreach to local students, and opportunities to identify strong candidates for internships and research positions.

What is the benefit to the community?

The hackathon strengthens the global microscopy + ML ecosystem by producing open-source code, curated datasets, tutorials, and new ideas. It promotes cross-institution collaboration, connects microscopy scientists with the mainstream ML community, and encourages reproducible, shared tools that others can build on.

What happens at a local site during the hackathon?

Local sites host group work sessions, discussions, and social time (during lunch).

Participants work together in teams, use shared resources, and stay connected to the global event through Slack and the main program. Sites can also run their own micro-activities — troubleshooting, brainstorming, short walkthroughs, etc.

Who can participate?

Anyone with interest in ML, microscopy, imaging, or materials science:

  • Undergraduate students
  • Graduate students
  • Postdocs
  • Faculty
  • Industry researchers
  • Independent learners

No prior microscopy experience is required.

But what if I am not familiar with ML/Python?

As long as you understand your problem from microscopy side, code assistants like ChatGPT and teaming up with the ML experts can be a way to proceed! Hackathon is also the environment to build interdisciplinary teams.

Should the data and code be open?

Yes. All submitted data, code, and results must be fully open.

The goal of the hackathon is to create a shared ecosystem of tools, datasets, and workflows that the entire community can reuse and build upon. Open data ensures reproducibility, enables follow-up collaborations, and allows future participants to extend and improve the submitted projects. Openness is a core principle of this event.

Who supports the awards, and how will they be paid?

Awards are provided by sponsors such as Renaissance Philanthropy, Covalent Metrology, Thermo Fisher Scientific, Theia Scientific, MSA Student Council, Toyota Research Institute, Waviks, Polaron and others.

Awards may be paid directly by the sponsor in cognizance of possible political and other restrictions. Some sponsors may choose winners independently, while others committed to follow jury recommendations.

Can we provide our own data for participants?

Yes. Additional datasets are welcome.

Organizers ask for advance notice because datasets must be converted into the digital twin microscope–compatible format, documented, and added to the central GitHub repository.

Rama Vasudevan (vasudevanrk@ornl.gov) and the core team will assist with formatting and integration.

What level of coding experience is needed to participate?

Basic Python is sufficient.

We provide starter notebooks, data loaders, digital twin examples, and template workflows. More experienced coders can dive deeper into ML modeling, active learning, or real-time analysis, but beginners can still contribute meaningfully.

If you are new to Python, this article is a great place to start:
“The New Language of Science: How to Learn Python Effectively” — https://medium.com/@sergei2vk/the-new-language-of-science-how-to-learn-python-effectively-c8ce51012a64

How do teams form?

Teams can form:

  • Through the hackathon Slack channels (recommended)
  • By browsing project ideas on the Miro board and reaching out to organizers or participants
  • Also, at local sites

We encourage teams with mixed expertise (ML, microscopy, physics, coding, domain knowledge), but single-person teams are also allowed.

How is the hackathon organized?
  1. Pre-Hackathon Launch (≈2 weeks before): We introduce the problems and datasets, show where participants can communicate (Slack, local sites) to form teams, and explain how to access the digital twins.
  2. Main Hackathon (3 days): Opening session, mentoring, Slack support, hands-on work with digital twins and datasets, collaboration across local sites and online, and final project submission.

After the hackathon, organizers coordinate judging, feedback, and joint paper writing.