Greenhouse Canada

News
Technology Issues: Take notice: Unseen AI in the greenhouse


January 29, 2019
By Dr. Jason Behrmann

Rapid innovation in machinery and computer technology have lowered barriers to entry within automation. New cutting-edge tools, from robots to artificial intelligence (AI), capable of complex tasks are increasingly available – and affordable.

For greenhouse farming, however, discussions about automation tend to focus on robots and machinery, leaving little mention of stealth technologies that do important work behind the scenes.

Bringing AI into the limelight
A future where robots replace greenhouse workers is arguably premature and overlooks important greenhouse automation technologies on the market today. One noteworthy development is AI tools that serve as digital assistants, not replacements, for greenhouse workers. Contrary to popular thought, sophisticated smart software can now be trained on tasks in greenhouse farming using big data, which is a lot of digital information on daily greenhouse operations. It can include anything from records of your power consumption to digital photos of your crops. From this, we can train AI software to learn common, repetitive tasks, reinforcing the decision-making capacities of skilled employees. Rather than replace, these AI-automation tools empower and streamline daily operations.

Advertisment

Consider this: Using image data banks of diseased and infested crops, smart software can learn patterns in spots, scars and insects on plants. You can take a photo of a diseased crop and obtain a diagnosis of the problem, along with a recommended treatment strategy. All the complex computations that automate disease diagnoses are on an unseen computer server far away.

Additional examples arise from your growing conditions. Using records of your heating, lighting, humidity and more, AI software can understand multiple factors that influence the growth of your greenhouse plants. Once known, smart software can analyze current growing conditions and make accurate predictions about the future productivity of your greenhouse, including harvest yields and flowering time, while reducing the time you devote to manual, repetitive plant assessments. Once again, the underlying computations remain unseen and you don’t need an elaborate website to access the information. At its most basic form, you could receive the results in an email along the lines of “you will obtain X tomatoes this week” or “your kalanchoe will be ready for sale by date Y”.

We are just beginning to harness the full potential of AI. Gone will be the days of a guess-and-check approach. Future AI tools will predict how yield and growth of ornamentals will change if a grower would alter, say, CO2 levels and supplemental lighting. By crunching millions of data points on unseen servers, emerging AI technologies show potential in flagging growing conditions that cause skin cracking and blemishes in produce. The secret to growing the perfect tomato or pepper may thus reside within smart software rather than at the fingertips of a robot.

Setting New standards
With the growing availability of big data in greenhouse farming, automation enabled by AI will set new industry standards for efficiency. Expect advancements in areas ranging from fully automated grow protocols at the push of a button and better pricing of produce, to improved labour management and more efficient use of energy.

Technology titan Google reduced energy consumption by 40 per cent at one of their facilities once their AI software learned how best to operate cooling systems. This groundbreaking achievement required no major equipment or changes to the building’s infrastructure, just smart software able to predict ahead of time when a facility would heat up and by how much. Expect energy-saving innovations like this to become standard in diverse industries, including agriculture, and influence future policies for clean energy and greenhouse gas emissions.

Though they may lack the shine and satisfying mechanical hum of machinery, unseen AI tools could play an important role in automating greenhouse cultivation. The time is now for growers to embrace this reality. Technological progress can soon improve many tried-and-true processes in the greenhouse – or make them obsolete.


Jason Behrmann, PhD, is the senior marketing communications manager at Motorleaf. He can be reached at jason.behrmann@motorleaf.com


Print this page

Related



Leave a Reply

Your email address will not be published. Required fields are marked *

*