Active Food & Agriculture Engineering

Modular Immersive Crop System for Monitoring and Harvesting Via Gentle Manipulation

Funder EPSRC (part of UKRI)
Recipient organisation Swansea University
Funding Not disclosed
Period Mar 2025 — Mar 2028

In plain English

AI plain-English summary

A vertical farming system that uses artificial intelligence and robotic arms to monitor and pick crops could make year-round food production cheaper and more accessible in low-resource regions. Current vertical farms rely heavily on labour, electricity, and expensive equipment, which limits them to high-value, small-scale crops grown in wealthy countries. The Harvest project aims to change that by building a modular, low-maintenance unit that can grow large crops such as tomatoes or peppers in a fully automated way. Sensors—infrared, thermal, and RGB cameras—surround the plants to create a three-dimensional picture of their health, detecting disease, pests, or water stress without human inspection. A compliant gripper on a flexible robotic arm then harvests the fruit, using the sensor data to avoid damaging the plant or other fruit. If successful, the system could reduce the cost and expertise needed to run a vertical farm, making it viable in developing countries and for domestic users who want a single unit at home. It could also allow non-native crops to be grown locally year-round, shortening supply chains and reducing reliance on imports. The project does not address the energy costs of vertical farming directly, but its low-intervention design could lower electricity and resource use compared to existing systems.

View original technical description
Current vertical farming systems are aimed at producing small fruit and vegetable crops on a large scale, with little diversity, and minimal profit margins. The high costs of running a vertical farm are down to labour, electricity, and resources which means current vertical systems are less likely to be used in third world countries and areas in short supply. The Harvest project is aimed at producing a low maintenance, low cost solution allowing for round year production of crops with minimal intervention. The aim is to produce a modular system allowing to grow multiple large crops vertically, with an artificial intelligence-based system to monitor the crops health, and harvest. The health of a crop is monitored with a wide variety of sensors (including infrared, thermal, and RGB cameras). By surrounding the crops with these sensors, we aim to build an immersive, 3 dimensional analysis of the plants, with the ability to detect disease, pests and stress. The harvesting of the fruits will be done using a compliant gripper, fixed to a continuum robotic arm. This allows for gentle manipulation to harvest the fruit, and using the immersive technology, avoid any obstacles to prevent crop damage and disturbance. Since the system will be fully automated, there will be less of a requirement for human intervention and expertise, this not only allows for non-native crops to be grown and cared for on a large scale, but also extends to domestic users being able to acquire single units for personal use.

View the original record at the funder ↗

Researchers

Robert Whittey (Student)

Related Research

Grants with similar aims, by meaning.

Robotic system for harvesting assistance.
Robust robot location and behaviors for on-farm navigation.
Autonomous monitoring and control of crop growth as a feedback system
Robotic Courgette Harvester
MoBile Robotic PlAtforms for ACtive InspeCtion and Harvesting in AgricUltural AreaS

Original classification

Studentship

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.