Today's Edition
MUDHI ALOBTHANI (ABU DHABI)
A team of researchers at Khalifa University is developing an advanced AI-powered agricultural robot capable of autonomously inspecting crops, identifying plant diseases and carrying out precision spraying inside greenhouses - a breakthrough that could strengthen the UAE's food security and sustainable farming ambitions.
The innovation is being led by PhD candidate Muhammad Salman at the university's Agri Robotics Laboratory under the supervision of Professor Irfan Hussain, with support from UAE agricultural partner Silal.
Designed for the UAE's harsh climate and controlled-environment farming sector, the system combines robotics with agentic artificial intelligence, enabling the machine to make decisions, interpret natural-language instructions and adapt its actions in real time.
"The UAE's high temperatures and extreme climate impose real limits on traditional agriculture, which is why controlled-environment farming has become so central here," Salman said. "But even inside greenhouses, operations remain heavily labour-intensive and wasteful in terms of water, chemicals and energy."
Autonomous Inspection and Precision Spraying
He said the robot addresses one of the biggest challenges facing UAE greenhouse farming: early disease detection and targeted treatment.
"The UAE grows most of its fresh produce inside greenhouses, which are labour-intensive and vulnerable to diseases that spread quickly in warm, humid conditions," Salman said.
"My work addresses the autonomous inspection and precision spraying side of that problem, catching plant diseases early and treating only the affected plants instead of blanket spraying entire rows."
The project began in simulation before progressing to physical deployment. "We developed AGRI BOT, an agentic system where a large language model generates the robot's behaviour on the fly from a natural instruction, combined with a vision language model that identifies diseased plants," Salman said.
The system has since been integrated into a TiAGo PRO robotic platform and is currently being tested in a real greenhouse environment.
"Our disease detection models reach around 94% accuracy, and the language-driven planning side currently reaches about 82% task alignment on realistic greenhouse instructions," Salman said.
He noted that one of the key technical challenges was enabling large AI models to run directly on the robot in real time. "Memory, computing speed, and storage are all tight when a language model and a vision model need to work together in real time," he said.
A General-purpose Greenhouse
Assistant Professor Hussain said the project reflects the UAE's broader efforts to use advanced technology to tackle strategic national priorities.
"The UAE imports around 80% of its food, and after COVID-19, food security shifted from a background concern to a clear national priority," Hussain said. "In a region with 45°C summers, limited fresh water, and limited rainfall, robotics and AI are not optional technologies."
He said the innovation stands out because of its adaptability compared with conventional agricultural robots. "Most agricultural robots today are engineered for a single, narrow task," Hussain said.
"The significance of this work is that it gives a robot the reasoning and adaptability to handle multiple tasks from a simple natural-language instruction. It moves us closer to a general-purpose greenhouse assistant rather than a fixed tool."
The research is supported through Khalifa University's robotics facilities and industry collaboration with Silal, which provides access to real greenhouse environments for testing.
Food Security and Sustainability
Looking ahead, the team plans to commercialise the technology through both autonomous greenhouse robots and compact handheld disease-detection devices for agronomists.
Both researchers said the technology directly supports the UAE's food security and sustainability strategies by reducing pesticide use, conserving resources and increasing domestic food production.
"By detecting diseases earlier and spraying only where needed, we help growers save crops and cut pesticide use significantly," Salman said. "That directly supports the UAE's goal of producing more of its own food, with less water, less chemical input, and less manual labour."
Hussain added: "Technologies like this help the UAE grow more of its own food more sustainably. The goal is a farming system that is productive, resilient, and uniquely suited to this region."
