Swarm robotics is a new type of robotics that allows simple individual robots to work together to perform complex tasks. Swarm robots have potential to be used for search and rescue, micro medicine, assembling structures and many other applications. These robots are biomimics; their behaviour is inspired by nature, and copies how bees and ants work together in their colonies. A huge advantage of using swarm robotics is that the sum of the individuals’ capabilities can make the swarm much cleverer than a single, much more complex robot. In some scenarios several simple robots working together can perform tasks more quickly, effectively and at a lower cost than a single specialised robot.


This project looks at how swarm robots can be used for precision agriculture. New precision farming practices are far more efficient and waste fewer resources than conventional techniques. The aim of precision agriculture is to apply agrochemicals (fertilisers, pesticides, herbicides) to the areas where they are most needed, at a given time, instead of the traditional approach to spray whole fields uniformly almost every day. However, there is currently no cost-effective way to accurately monitor and manage crop health and chemical supply at an individual plant level.



Your task is to design a sustainable swarm robotics system that is able to perform the monitoring step for a precision agriculture application from the level of many fields down to the level of individual plants. The result should be a reduced but more effective use of agrochemicals to give a higher quality and quantity of product crop. You will need to thoroughly research applications of swarm robotics, and precision agriculture in order to develop both the robots and any supporting infrastructure.



Sensors and data fusion

Sensing is the key aspect for monitoring crops. You will need to consider what kind of information is required for precision agriculture and define which types of sensor are necessary to gather the required data. How will you ensure your sensors are accurate, reliable and giving complete data; will any calibration be required? Your system will need to process a lot of data from many sensors. In your design work you should consider data fusion and sensor fusion algorithms. Currently satellite imagery is used perform analysis of crops at the level of the field, and this data is used to predict crop yields[1].

One important issue in commercialising your robots is their flexibility. Pineapples are likely to be monitored differently from potatoes: can your robots adapt to all crops, or will they have a specific application? How will you convince investors of your chosen strategy of either a single-crop robot, or one which can monitor 6 crops and may have a much greater market?

[1] – Cropio is one of a several satellite crop monitoring systems.