Adeayo Adewumi, a Nigerian researcher with Purdue Research Services in the United States, is gaining global recognition for his groundbreaking work in agricultural innovation.
At the recent American Society of Agricultural and Biological Engineers (ASABE) conference in Toronto, Adewumi received multiple awards for his research focused on optimizing artificial intelligence (AI) algorithms for real-world application in agricultural fields.
In an interview, Adewumi discusses the significance of his research and its potential to transform farming practices both in Nigeria and globally.
He emphasised how local agricultural challenges can be addressed using global knowledge and cutting-edge technology like AI to boost productivity, efficiency, and sustainability.
Congratulations on your awards at the just concluded ASABE conference. Can you share your experience at the conference and your awards?
Thank you for the opportunity to discuss my work. ASABE is a leading professional and technical organization committed to advancing engineering applications in agricultural, food, and biological systems. Its Annual International Meeting (AIM) brings together researchers, engineers, students, industry professionals, and policymakers from around the world to share innovations and collaborate on global challenges.
At AIM 2025, I presented my research as a member of the organization, made valuable professional connections, and participated in research competitions. I was honored to receive second place in the Research Awards from both the African Network Group of ASABE (ANGASABE) and the Association of Agricultural, Biological, and Food Engineers of Indian Origin (AABFEIO). While my affiliation with AABFEIO may seem unexpected, I am a member of the group through my advisor, Dr. Dharmendra Saraswat, whose heritage connects him to the community. These awards recognize excellence in the conduct and presentation of research in agricultural, food, and biological engineering.
Your research explores artificial intelligence and edge computing for agricultural applications. What inspired this work, and what are its key objectives.?
My inspiration stems from the intersection of computer vision, deep learning – a subsect of Artificial Intelligence – and edge computing for agricultural applications. Deep learning methodologies are advancing how computers identify and locate elements within agricultural landscapes, from livestock to crops, weeds, and diseases. For agricultural systems operating in remote areas with limited internet connectivity and resources, edge devices performing real-time detections are essential. This is important for different agricultural tasks such as phenotyping, weed and disease identification and mitigation, precise nutrient application, yield prediction, and crop harvesting.
The primary challenge we address is that when deep learning detection models are deployed on resource constrained edge devices some metrics such as inference time (speed at which these models run) often suffer. One way to speed up these models is by using an optimization method called TensoRT but while TensorRT optimization has been used to enhance detection performance on edge devices, there was a knowledge gap regarding how this optimization affects metrics beyond inference speed, particularly detection precision. Our objective was to evaluate the impact of TensorRT on the precision performance of selected models across two edge devices.
What are your main findings, and how are they particularly relevant to the needs of countries like Nigeria?
The main takeaway from the study is that while high-cost edge devices improve inference speed, they do not maintain the precision performance of deep learning algorithms when compared to the lower-cost device used. My goal is to see this research directly translate into practical, impactful solutions for farmers and agricultural systems globally, especially in regions like Nigeria where technological advancements can yield profound benefits.
Our findings guide the deployment of optimized detection systems on resource-constrained edge devices for agricultural applications. By enabling highly efficient, real-time object detection on compact, energy-efficient devices, this research helps to balance detection accuracy with computational efficiency in bandwidth-limited, battery-constrained environments. This means farmers can implement sophisticated AI-powered tools directly in the field for automated agricultural tasks.
What are the current challenges and advice to policy makers on pathways to deploying AI in Nigeria’s agricultural systems?
There are different efforts from the public, private and international organizations to accelerate the AI deployment in food security. For instance, the Minister of Communications, Innovation, and Digital Economy, Dr. Bosun Tijani, has a good knowledge of the AI space and recently emphasized that AI is central to boosting productivity in agriculture, public health, and beyond. In addition, we have seen collaboration with the Gates Foundation to provide resources, mentorship, and support necessary to translate AI ideas into impact in agriculture, health and education. There are also other capacity building initiatives via the National Information Technology Development Agency (NITDA).
Despite growing efforts, some barriers limit the impact of AI in farming. Rural infrastructure gaps, such as unreliable internet and inadequate power supply, hinder the deployment of cloud-based and sensor-driven solutions. Low digital literacy among smallholder farmers also slows adoption, as the benefits of AI can only be realized when users are equipped to leverage the technology. Additionally, the high cost of advanced tools like drones, IoT sensors, and satellite imaging makes them inaccessible to many small-scale farmers who form the backbone of the agricultural sector.
To enable effective deployment of AI in Nigeria’s agricultural sector, the policy and innovation community investments in rural broadband and electricity infrastructure are laying the groundwork for tech-driven farming. Public-private partnerships to bridge funding and expertise gaps, and equip farmers with smart devices and digital advisory tools.
Curriculum reforms in higher education and targeted rural training to build the skills required for a digitized agricultural workforce. Additionally, programs aimed at subsidizing AI tools, promoting local manufacturing, and expanding agri-tech platforms will make advanced technologies more affordable and accessible to smallholder farmers.
For young Nigerians passionate about agricultural technology, what advice would you offer as they pursue careers that can drive innovation in their home country?
My advice is to seek global experiences that expand your knowledge and network but always look for ways to apply those skills to address local challenges. Nigeria is full of potential, and there’s a significant need for skilled professionals to drive innovation and development, particularly in sectors as vital as agriculture.
Whether it’s through research, entrepreneurship, or public service, every contribution counts. Focus on understanding the specific needs and constraints of your local environment and then adapt the global knowledge and technologies you acquire to create sustainable, impactful solutions. The future of Nigerian agriculture will heavily rely on locally driven technological advancements.
