Photonics on the farm: robotics to help feed the world

Simon Blackmore talks about farming with robots
for precision agriculture in an
SPIE Newsroom video interview [6:58].

Ten to 15 years ago, farmers used to laugh when Simon Blackmore and his colleagues talked about deploying robotics for such chores as weeding, protecting crops from disease or pests, or selecting harvest-ready vegetables — all while helping to cut costs and limit chemical and other impacts on the soil.

Now, he said in an SPIE Newsroom video interview posted last week, they’re asking questions about how robotics and other photonics-enabled technologies can help save energy and money, minimize soil damage, and improve crop yield.

Blackmore, who is Head of Engineering at Harper Adams University in Shropshire, director of the UK National Centre for Precision Farming (NCPF), and project manager of FutureFarm, also shared his ideas in a new conference at SPIE Defense and Commercial Sensing in April on technologies with applications in precision agriculture.

Blackmore and his NCPF colleagues are working to overhaul current farming practices by intelligently targeting inputs and energy usage. Their lightweight robots are capable of planting seeds in fields even at full moisture capacity, replacing heavy tractors that compact and damage the soil.

Simon Blackmore

Robots have also been designed with micro-tillage capabilities, to target the soil at individual seed positions, and for selective harvesting of crops for quality assurance.

“Now one of my former PhD students has developed a laser weeding system that probably uses the minimum amount of energy to kill weeds, by using machine vision to recognize the species, biomass, leaf area, and position of the meristem, or growing point,” Blackmore said.

A miniature spray boom of only a few centimeters wide can then apply a microdot of herbicide directly onto the leaf of the weed, thus saving 99.9% by volume of spray. Or, a steerable 5W laser can heat the meristem until the cells rupture and the weed becomes dormant. These devices could be carried on a small robot no bigger than an office desk and work 24/7 without damaging the soil or crop.

Not surprisingly, data is a hot topic in the field of precision agriculture.

Several speakers at the April event — among them John Valasek, and Alex Thomasson of Texas A&M University (TAMU), chairs of the conference, and Elizabeth Bondi of the Rochester Institute of Technology (RIT) — spoke about best practices for collecting data, and Kern Ding of California State Polytechnic University discussed data processing techniques.

Valasek also described several sensors and different ways they may be flown. Factors such as weather, speed, altitude, and frame rate can dramatically change the quality of the data products from UAV imagery.

Bondi discussed the calibration of imagery from UAVs (unmanned aerial vehicles, such as drones) to maintain consistency over time and under different illumination conditions.

Other speakers — Haly Neely of TAMU, Carlos Zuniga of Washington State University, and Raymond Hunt of the U.S. Agricultural Research Service — focused on the use of UAVs for such applications as soil variability, irrigation efficiency, insect infestation, and nitrogen management for crops including cotton, grapes, and potatoes.

Plant phenotyping — the analysis of crop characteristics such as growth, height, disease resistance, nutrient levels, and yield — is vital to increase crop production. Taking these data with current methods can damage plants, and is time-consuming and expensive. UAVs, carrying the right sensors, have the potential to make phenotyping more efficient and less damaging.

Speakers Yu Jiang of the University of Georgia, Andrew French of the U.S. Arid-Land Agriculture Research Center, and Grant Anderson of RIT described ground-based systems to expedite phenotyping, and Joe Mari Maja of Clemson University, Yeyin Shi of TAMU, Maria Balota of Virginia Polytechnic Institute, and Lav Khot of Washington State University discussed UAV-based systems.

With images and measurements from such devices, for example, cotton height may be determined and cotton bolls counted, soil temperature can be mapped, and nutrient levels in wine grapes were assessed remotely.

Small- and mid-sized farms are expected to see the largest yield increase from these initiatives. The ultimate result of all this photonics-enabled precision agriculture is profound: healthier food, more productive farms and gardens, and more nutritious food for a growing world population.

Thanks to Elizabeth Bondi and Emily Berkson, both of RIT, for contributions to this post.