Dr. Mike Duncan: Changing the face of ag tech

Inside the research lab at Niagara College’s Agriculture & Environmental Technologies Innovation Centre where Mike Duncan,PhD and his team are exploring sophisticated technologies, including robotics and artificial intelligence (AI) techniques.

Without question, Carl Sagan had an effect on many scientists during his life and enduring legacy. For Research & Innovation’s Mike Duncan, who has his PhD in Agricultural Physics, it was a story by the famed astronomer that intrigued him years ago. Sagan made the comparison between people and snails and pondered how long it would take people to evolve to the point where they too could carry a house around on their backs.

The answer, of course, is that humans developed a brain – an evolution that has equipped them to invent technology and build their own houses, explains Duncan, adding that human innovations have allowed them to adapt faster and more easily to any environment on earth than their snail counterpart, who spent 500 million years perfecting its house.

“The point is that we can very quickly adapt and evolve techniques and tools that can solve most of the problems we have,” says Duncan, who makes it his mission at Niagara College to examine, and define the future technologies that will change the face of agriculture.

He recognizes the severity of challenges currently facing farmers – a rising global population, with an urgent need to produce more food, more efficiently and with more environmental stewardship – in the face of fewer farmers, a shrinking amount of farmed land, and climate change realities. Still, he believes that precision and regenerative agriculture technology may just help save the planet.

“We are very much capable of mitigating all these problems through technology and the technology is a by-product of our own evolution,” he insists. “This is true even if all the technology does is tell us how many trees to plant and where to plant them.”

With an interdisciplinary background that includes academic pursuits in atmospheric remote sensing using aircraft and radar, as well as business experience in software development, supercomputers, and large-scale virtual reality, Duncan has found creative applications for all these skills since arriving at the College in 2001 as an industry expert in virtual reality. He was a founder of the Centre of Advanced Visualization (CFAV), a research group dedicated to exploring the use of virtual reality for urban and land use visualization. He has been the recipient of major awards and grants both provincially and federally, which allowed him to found the Augmented Reality Research Centre (ARCC) in 2006 to expand his research into precision agriculture, agricultural remote sensing, and visualization.

Today, Duncan is recognized as a world leader in precision agriculture – which utilizes smart technologies aimed at using fewer resources to grow more – and is entering his eighth year as the Natural Sciences and Engineering Research Council’s Industrial Research Chair for Colleges (NSERC-IRCC) in Precision Agriculture & Environmental Technologies at the College.  

“We are very much capable of mitigating all these problems through technology and the technology is a by-product of our own evolution.”

He’s a devoted proponent of transparent knowledge transfer in the pursuit of helping the planet and partners within farming communities throughout Ontario and Canada while working with his research team, comprised of computer programming, electronics, robotics, and GIS students, and recent graduates. With the appointment of the national research chair came the founding of the Agriculture & Environmental Technologies Innovation Centre (AETIC) at Niagara College. The centre carries on the agriculture technology work he first started and collaborates with a wide variety of partners, including farmers, commodity organizations, agricultural consultants/agronomists, agricultural technology organizations, and food companies. 

In the early days of AETIC, Duncan was the principal designer of Niagara College’s Crop Portal, an interactive web software system that houses and processes farm data, such as yield and topography, into colour-coded maps, allowing farmers and crop consultants detailed insight into their fields’ productivity variability. The platform has recently been expanded to allow farmers and scientists not only more flexibility to visualize and verify their data, but to also have the capability to manipulate their own algorithms.

Niagara College’s Crop Portal is an interactive web software system that allows farmers and crop consultants detailed insight into their fields’ productivity variability.

 

And more recently, he has dug deep into new and sophisticated technologies, including robotics and artificial intelligence (AI) techniques. The AI work focuses on the ‘suitability problem’ while the robotics is aimed at in-field 24/7 remote sensing work. 

“The suitability work partially addresses climate change concerns where weather data is used to assess the viability of crops in different areas as the weather changes,” explains Duncan, “and the robotics work will initially address autonomous navigation and then move on to problems such as estimating grape yields in vineyards.”

For these pioneering endeavours, Duncan has brought together agricultural companies, agTech investors, tech-savvy farmers, and inventors for agri-food and agriculture technology collaborations in applied research at the College.

“The idea is to both examine, and define, the future technologies that will drive a remotely-operated farm business,” he says, adding that for project partners, this will create intellectual property to refine current and develop new products and services.

One of AETIC’s most ambitious and far-reaching projects is the work being done with Italian confectioner Ferrero. The research project involves the idea of the suitability of an area to a hazelnut crop and is applicable globally and under a variety of climate change scenarios. The area where a crop grows is characterized by very detailed weather data and that characterization is used by seven AI algorithms to compare with other potential growth areas for that crop. 

“The idea is to both examine, and define the future technologies that will drive a remotely-operated farm business.”

The detailed weather data is generated using a numerical weather model, which can re-create the weather over the last 18 years in Southern Ontario at one-hour intervals. Indeed, the model generates 140 weather variables at over 30,000 points, to allow the AI algorithms to make very accurate maps of suitable growing areas for a given crop. 

“It also allows us to evaluate whether crops that currently grow in the area will be able to grow given various warming scenarios resulting from climate change.” 

The technique being developed is very flexible, and the weather data, which is being generated by the wizards at the environmental engineering firm RWDI, is also applicable to a huge variety of problems, past, present and future, which Duncan points out was the reason for getting involved in the project in the first place.

“The ‘forensic’ weather data can be used to go back and characterize the growing conditions for any crop in Southern Ontario in the last 18 years.”

The agricultural resources at Niagara College include a 40-acre campus vineyard, which Duncan first instrumented with multiple temperature sensors in late 2006. Sixteen sensors revealed a virtual cacophony of variability. Duncan remembers seeing and recording temperature changes of 10 degrees over distances as short as a few metres and over a few minutes. However, when he consulted with experts, he was told that his sensors were ‘broken.’

“At that point, I realized that I might be doing something new,” he says. No one in the field was used to looking at the really detailed behaviours of temperature, and as it turns out, most other variables as well. He has since developed a unique set of analytics that can deal with this variability and provide farmers with better information.

The research lab’s development of agricultural robots will help the AETIC team develop a stable, self-navigating platform that can then act as a prototyping tool for industry partners to conduct their own research. The intent of this work, explains Duncan, is to develop a generally useful platform that can then be applied to a large number of farms and other challenges.

The newest member to the AETIC team is an advanced, rugged land rover, called RoamIO Jumbo, which is able to patrol vineyard rows with ease 24/7 thanks to an onboard generator. The land robot, built by Korechi Innovations, will be able to carry a number of sensors, including thermal imagers, traditional image capturing cameras, LiDARs, SONARS, temperature probes, relative humidity probes, soil moisture probes, and more. All these sensors will help farmers increase profitability or even help save their crops. And with this innovative platform, there is potential to incorporate AI to analyze the data in real-time.

Dr. Mike Duncan with the RoamIO Jumbo, a state-of-the-art agriculture robot capable of carrying a number of sensors to help farmers increase profitability or even save their crops.

With the grape growers in Niagara, Duncan’s current work with Korechi robotics includes the collection of crop imagery in vineyards for grape health analysis, ripeness estimation, and ultimately yield prediction. And for its work with Ferrero, the state-of-the-art land rover will be able to sense nut grove conditions in real time and can even be mounted with an industrial fan to help with pollination of hazelnut trees should a lack of wind in the tight growing season occur.

The advantage of ground drones, such as RoamIO, over conventional air drones, is that the cameras mounted on it are close to the ground, and in the case of grape vines, this allows them to look up into the canopy, giving access to parts of the vines where diseases and infestations occur. The same is true in orchards and plantations.

There are very few robots out there in farm fields to take on these complex challenges, so Ontario may soon find itself at the forefront of agriculture robotics technology, adds Duncan. Precision agriculture is the future of farm businesses and Niagara College is taking a leading role to help growers reach profitability, efficiency, and sustainability on the farm.

To learn more about the work of the Agriculture & Environmental Technologies division, visit the web page.