About AgriTrait

Before AI models could begin translating multimodal plant signals into near-continuous fertigation decisions, a scientific foundation first had to be established across materials science, sensing fidelity, and plant–environment control systems. That foundation was laid through the Phase 1 iSURF programme (Intra-CREATE funded project), co-led by Prof Ng Kee Woei, Chair of NTU’s School of Materials Science and Engineering (MSE). The work advanced a new research paradigm by shifting from predetermined ambient passive experimentation towards using momentary whole-plant physiological status as the primary reference point for decision-making.

Building on this foundation, our project, AI Model-as-a-Service for Data-to-Decision Agriculture (Agri-TrAIt), develops the AI layer needed to translate rich agricultural data into actionable decisions. AgriTrait aims to develop methods to identify low-cost proxy measurements for key crop traits, learn adaptive intervention strategies to improve yield and nutritional quality, and enable knowledge transfer across farms under practical constraints such as limited sensing and privacy.

The project is supported under the National Research Foundation (NRF)’s AI for Science Catalytic Grant Call, and brings together expertise from both the science and computing communities. A video introduction of AgriTrait is below. Note that some information is outdated as it was prepared in September 2025.