Data-smarter farming: AI-blockchain system to boost crop health
New CQUniversity research has taken the battle against crop disease into the algorithm, developing a model for smart agriculture systems through artificial intelligence (AI).
The innovation gives farmers a smart way to boost productivity, while ensuring crop management is transparent, secure and traceable across the food supply chain through blockchain technology.
Dr Gopi Akella developed the framework as part of his doctoral research at CQUniversity, and graduated with his PhD at the CQU Melbourne graduation ceremony on Wednesday 12 August.
His research, Convergence of AI and Blockchain: A Modular Architecture for Smart Agriculture Systems, combines advanced AI analytics with blockchain-enabled record keeping to improve decision-making, traceability and transparency throughout the food-production lifecycle.
Tracking agriculture
Dr Akella is an experienced project manager and Information and Communications Technology lecturer, and said his research tackled a gap in how farmers record and verify their decision-making.
“Agriculture generates vast amounts of data from in-crop sensors such as weather and market conditions, but farmers need to be able to record and track the patterns, and their own decisions and inputs,” Dr Akella explained.
“There has been limited practical supports for farmers to record and verify their decisions using technology.
“Smart agriculture is promoting sustainable farming practices and global food security, and AI and blockchain can unlock more efficient ways to trace and share best-practice farming, from farm to table.”
Dynamic shift
At the heart of the research is Dr Akella’s new AI approach called the Dynamic Self-Attention (DSA) algorithm.
Dr Akella said developing a new approach required careful examination of the limitations of existing deep-learning and attention-based methods, while creating a solution that was both lightweight and interpretable.
He explained that many AI models operate as 'black boxes', with hidden processes from input to output.
“The DSA algorithm is designed to remain responsive to changing inputs while offering greater interpretability, creating opportunities for more transparent agricultural decision-making,” he said.
Dr Akella hopes the research will now progress beyond theory and into real-world implementation.
He believes the architecture could support the development of next-generation smart agriculture systems capable of improving automation, strengthening trust across supply chains and helping producers make more informed decisions.
For Dr Akella, graduating from CQUniversity marks the culmination of years of research and the beginning of new opportunities to translate emerging technologies into practical solutions in the vital agricultural industry.
Dr Akella’s supervisors were Associate Professor Santoso Wibowo, Assoc Prof Srimannarayana Grandhi, Dr Sameera Mubarak and Dr Fariza Sabrina.
Learn more about Higher Degrees by Research at CQUniversity.
