It’s farming, but not as we know it

We’re all familiar with the challenges of global agricultural production – how to increase production while lowering costs, overcoming climate changes and optimizing the supply chain. Vision is now playing a huge role in addressing these challenges and, along with AI and increased data processing, offers a range of solutions to today’s farmers and retailers.

Agriculture 4.0, Farming 4.0, AgriTech and Smart Farming are all recognized terms being used to refer to the adoption of advanced technology to improve yields and feed the world. These follow on from movements such as Mechanized Agriculture at the turn of the 18th century and the Green Revolution commencing in the 1960s.

Precision Agriculture (PA) is a model which is enjoying wide adoption. This concept relies on using an array of sensors and collection of vast amounts of data to improve yield and reduce waste. Several elements of PA rely on vision to bring this concept to life.

Smart machinery is the new farm hand

John Deere made an impression at this year’s CES with its self-driving tractor and smart harvester. Using vision and GPS technology, the tractor will cover every inch of a field evenly and effectively, gathering data on soil conditions and crop quality. The giant combine harvester relies on computer vision and machine learning to actually “see” the grain in real-time as it is being harvested and improve the quality of the harvest. It’s clear now what John Deere had in mind when buying AI tech company Blue River Technology in 2017. Technology exhibited here, and in use by other companies globally, is helping to reduce the amount of pesticide used in the environment, optimise fertilizing and watering, sow more effectively and generally manage crop productivity.

On a greatly reduced scale, Small Robot Company (SRC) has had plenty of media coverage recently, introducing their family – Tom, Dick and Harry. These diminutive robots have been created by a partnership of farming and engineering know-how and will be leased out to farmers wishing to maximize yield while minimizing financial and physical waste. Tom’s part is to roll gently through the fields, digitizing data plant by plant and uploading his findings to help create a plan of what to do where. Dick is summoned to act on the data provided and will spray just the right amount of fertilizer or weed killer exactly where needed, or even remove weeds without chemicals. Harry is the sowing machine and will drill, place and record every seed individually and accurately without the damage that a large tractor or plough inevitably causes. The brains behind the robots is Wilma, the software system which processes image and sensor data, makes decisions and generates instructions based on the masses of data collected.

Of course, SRC aren’t the only ones embracing vision and robotics, the robot farming army is growing. Sweeper is a robot using 3D vision and object identification software to pick peppers. It was created by a European consortium as part of the Horizon 2020 program. In the US, Harvest CROO is introducing its strawberry picker to meet the challenge of a declining workforce and growing population. Its technology uses multiple robotic components and visual inspection to safely pick and pack produce.

According to a report conducted by Oliver Wyman, “by 2050 we will need to produce 70% more food”[1] and vision-guided robotics are certainly helping in the race to achieve this.

Deep learning improves sorting and categorization

Several of our news posts have looked at the role of machine vision and deep learning in high-speed inspection. It’s not a new concept, but it is becoming more universally accepted as a “norm” in food processing and manufacturing. RSIP’s ‘Computer Vision News’ January edition carried an interesting case study of using deep learning to grade fruit. RSIP partnered with Sunkist RTS to create a real-time inspection process where defects to the stem or blossom of the fruit do not result in downgrading an otherwise acceptable item.

The war of the cucumbers

Tech giants Microsoft emerged as winners as they faced-off with Intel, Tencent and a team of human horticulturalists to win a competition to grow the greatest number of viable cucumbers, sustainably, in an autonomous greenhouse in the Netherlands. A report here explains how the human team came in second amongst the AI challengers.

Machine vision, advanced image processing and the adoption of AI are coming together to change global farming methods. Don’t be left behind – view our product range and get in touch to see how we can help modernize your vision system and sign up to our newsletter to stay abreast of industry developments.

[1] https://www.oliverwyman.com/our-expertise/insights/2018/feb/agriculture-4-0–the-future-of-farming-technology.html