IAEA Launches Global Research Project to Predict Radiation Effects on Polymers
The International Atomic Energy Agency (IAEA) has announced a five-year research initiative asking scientific institutions to develop machine learning tools that predict how polymers behave when exposed to ionizing radiation. The project runs from 2026 to 2031 and aims to replace expensive trial-and-error testing with data-driven modeling across nuclear energy, healthcare, and manufacturing sectors.
Polymers-used in cable insulation, medical devices, and industrial components-undergo structural changes under radiation exposure. These include cross-linking that can strengthen materials, chain scission that weakens them, and oxidation that affects durability. Engineers currently must conduct time-consuming physical experiments for each new application, slowing development and raising costs.
The Data Problem
Machine learning has transformed weather forecasting and finance, but its application to radiation-polymer interactions has stalled due to fragmented data. Decades of research sit scattered across academic journals. Industrial data remains proprietary and inaccessible. No standardized, validated datasets exist.
This fragmentation prevents machine learning systems from identifying patterns needed for reliable predictions.
IAEA's Three-Part Approach
The Coordinated Research Project will build the first comprehensive global database of radiation-induced polymer changes through three pillars:
- Consolidating and standardizing existing research data while validating findings from decades of studies
- Conducting targeted experiments to fill critical gaps where evidence is missing or inconsistent
- Training AI systems to predict polymer behavior under radiation and simulate outcomes across different materials and conditions
The goal is to enable predictive modeling that guides polymer design without extensive physical testing.
Industry Applications
Better prediction tools could improve durability of reactor materials and cables in nuclear plants. Healthcare providers could optimize sterilization processes for medical equipment. Manufacturers could design more resilient plastics and composites. Environmental benefits would include reduced waste and energy use in material testing.
How to Participate
The IAEA is inviting research organizations worldwide to submit proposals. The agency encourages participation from women and early-career researchers and partnerships between academic institutions and industry.
Proposals are due by 29 May 2026 and should be submitted via email to the IAEA's Research Contracts Administration Section. Templates are available through the Coordinated Research Activities platform. Both research and technical contract proposals are eligible.
Organizations assigned to the project will focus on specific polymers and research tasks as part of the coordinated global effort.
What This Means for Research
The project combines nuclear science with machine learning to shift materials engineering toward data-driven prediction. If successful, it could reduce development timelines, lower costs, and enable more precise design of radiation-exposed materials. Industries would move from experimentation to prediction.
For researchers interested in how AI applies to scientific fields, AI for Science & Research covers data modeling and laboratory optimization-core topics of this initiative.
Your membership also unlocks:
