The camera-based system mounts on a sprayer boom and captures field conditions as farmers make passes across the field. Using machine learning, it analyzes thousands of images to generate crop and weed maps, along with plant stand counts.

Why it Matters: In-season imaging can give farmers a clearer picture of how crops are performing across different parts of the field, helping fine-tune future decisions.

The system was first introduced in a limited rollout in 2022 and went through on-farm testing before expanding more broadly. Early versions were installed on a small number of farms to gather data and refine the technology.

“We spend a couple of years using the product in the field before it’s commercialized,” company CEO Cory Willness said.

“By the time something gets commercialized, it has already been used internally and has been through pretty rigorous testing.”

Adding in-season data to zone maps

SWAT Cam is designed to complement the company’s core SWAT Maps platform, which divides fields into management zones based on soil, water and topography. While those maps are relatively static, SWAT Cam adds a layer of in-season observation.

As the sprayer moves through the field, cameras mounted on the boom capture images every 50 to 60 feet. Those images are processed to measure plant populations and identify weed pressure and are then overlaid onto existing zone maps.

“It’s really a tool that uses the SWAT Map as the base underneath to define what’s happening in different areas of the field,” he said.

The system is not designed to make real-time decisions. Instead, it provides information that farmers and agronomists can use to evaluate performance and adjust management in future seasons.

For example, growers can use the data to compare expected and actual plant stands or identify patterns in crop performance across different parts of a field. In some cases, that can lead to changes in seeding rates or other input decisions.

“It’s information that helps them make better decisions in the future,” he said.

“It’s like a quick check-up tool.”

Adoption to date has been concentrated in Western Canada, where the company operates both directly with farmers and through a partner network.

The technology can be used on crops such as canola, wheat and soybeans, although its effectiveness depends on crop stage and canopy conditions.

The milestone offers a snapshot of how Croptimistic’s platform is expanding beyond its original focus on soil-based mapping.

In recent months, the company has introduced a series of new initiatives aimed at building out that platform, including SWAT Labs, an in-house soil testing facility, and a white paper outlining its approach to regenerative agriculture.

Together, those efforts point to two sides of the company’s development — expanding its core services while building new tools and frameworks around them.

“We have what I call an innovation engine and a business engine,” he said.

SWAT Labs is an example of the “business engine,” reflecting an effort to bring more of the soil data pipeline under one roof, from sampling through analysis.

The white paper, meanwhile, falls under “innovation,” and signals a push to shape how that data is interpreted and used in agronomic decision-making.

“These initiatives are building toward bigger things,” he said.

However, as technology advances, one long-standing issue is coming into sharper focus: the data itself isn’t always consistent.

At the World Agri-Tech Global conference in San Francisco, one speaker pointed to a fundamental challenge facing soil innovation.

“We hear from startups that will send a sample to multiple labs and get wildly different results, even though it’s the same unit area,” said Sam Malloy, research and development director with the National Science Foundation Ascend Engine.

“And that’s a real challenge.”

Why it Matters: Better soil data can improve decisions, but variability in how samples are collected and analyzed can limit how useful that information really is.

That kind of variability makes it difficult to compare results, build large datasets or scale digital tools that depend on consistent inputs.

Yet even within that same event, speakers acknowledged that soil systems are complex and dynamic, and that perfect consistency may be difficult to achieve.

University of Saskatchewan soil fertility professor Jeff Schoenau agrees. While appealing, large-scale standardization simply isn’t practical.

“I can understand why they want to do it because when dealing with really big data sets, it’s easier to roll through all of that if it’s all standardized and consistent,” he said.

“But unfortunately, that may not be particularly biologically meaningful.”

Part of the challenge is that soil testing isn’t a single, uniform process.

Methods can vary depending on what’s being measured, from nutrient levels to organic matter or biological activity. Even within a single field, factors such as soil type, moisture and landscape position can influence results, making it difficult to apply one standardized approach across the board.

Sampling itself also introduces variability.

Differences in depth, timing and handling can all affect the outcome, and those decisions are often made with a specific purpose in mind, whether that’s fertility recommendations, environmental monitoring or research.

The risk of standardization is that it can strip away information that is important for interpreting how soil functions in real-world conditions.

That doesn’t mean consistency has no role, Schoenau said, but it has to be balanced with an understanding of how soil behaves as living systems.

Trying to force uniformity across all situations can oversimplify that complexity, particularly when biological processes are involved.

“You’d be kind of trying to lump everything into into one bucket,” he said.

“That bucket may not be the best fit for everything.”

For companies building digital tools around soil data, that variability presents a different kind of challenge: how to manage it without losing meaning. That’s the approach taken by Saskatchewan-based Croptimistic Technology.

The company’s Swat Maps system is built around the simple idea that production variability within a field can be explained, mapped and managed.

The process starts by layering soil, water and topography data to divide fields into distinct management zones. Those zones are then checked with soil sampling and agronomic knowledge to understand what’s actually driving differences in yield potential.

Rather than treating a field as a uniform unit, the system breaks it into consistent landscape positions, from dry hilltops to low-lying areas where water and nutrients accumulate.

That zoned approach allows farmers to vary seed and fertilizer rates based on the productive capacity of each part of the field, with the goal of improving overall performance rather than simply reducing inputs.

Working with variability, not against it

That focus on understanding variability, rather than smoothing it out, aligns closely with the concerns raised by soil scientists in San Francisco, but it also introduces another layer of complexity when it comes to the data itself.

Croptimistic soil scientist Joel Ens says that variability is something his team deals with directly.

Differences in extraction methods, sampling strategies and lab procedures make it difficult to directly compare results, particularly when data is being combined across multiple fields or regions.

“We can convert one lab’s results into another,” he said. “But there’s always that loss in accuracy.”

That’s part of the reason Croptimistic has moved to bring soil testing in-house with the launch of Swat Labs in January, a new facility designed to connect sampling, analysis and digital records within a single system.

By controlling that entire workflow, the company aims to improve consistency in how soil data is generated and interpreted. It also allows for the development of larger, internally consistent datasets that can be used to refine recommendations over time.

“We have direct comparable results to other fields,” Ens said.

“It’s not just what’s happening in your field.”

That broader dataset is central to how the system works. Rather than relying solely on individual soil tests, Croptimistic’s approach connects fields through shared data, allowing patterns to emerge across similar soil types and conditions.

At the same time, the company is careful not to treat that data as interchangeable across all situations. Comparisons are made within similar environments, recognizing that variability between regions and soil types still matters.

The goal isn’t to eliminate complexity, Ens said, but to manage it.

“When we bring everything together — mapping, sampling, prescription development — we can build tools so that complicated systems become easy to use,” he said.

Why it Matters: Emerging sensing and AI technologies could eventually improve soil analysis, equipment uptime and precision decision-making on farms.

Launched in 2019, the program pairs Deere with hand-selected startups for year-long, project-based collaborations designed to test how emerging technologies perform in agricultural and construction use cases. It is not primarily an acquisition or investment vehicle.

“We’ve intentionally designed it that way,” said Colton Salyards, who manages the program within Deere’s corporate development and strategy group.

“The program was never designed to be an investment or an acquisition vehicle.”

Instead, Deere and each startup define a joint project, outlining objectives on both sides and evaluating how a given technology might perform in agricultural or construction use cases.

With the addition of this year’s five companies, Deere will have worked with 42 startups through the program.

Each year, Salyards said, public announcements of the cohort generate significant inbound interest from startups hoping to participate. The response can be “overwhelming,” but the companies selected stand out.

“There’s a key reason why we’ve selected them,” he said.

“There are use cases across agriculture that we believe could be of tremendous customer value.”

Sensor sensibility

Among the 2026 cohort is Australian firm resonAg, which is adapting miniaturized MRI-based sensing technology-technology originally developed for medical imaging and later adapted for industries such as mining and oil and gas-for use in advanced soil sensing.

Deere is exploring how that sensing capability could support precision agriculture applications.

“This is of huge importance for precision agriculture,” he said.

“Imagine a planting system that can sense and act in real time to conditions across the field.”

Another company, AIRS ML, is developing edge-AI systems that combine machine sensor data with on-device machine learning to predict equipment failures in real time. The goal is to improve uptime by identifying potential maintenance issues before they lead to breakdowns.

The remaining companies in the cohort include:

• IoTag, which focuses on telematics and mixed-fleet performance insights.
• TorqueAGI, which is developing AI foundation models for robotics.
• Aerobotics, which applies drone imagery and computer vision to specialty crop production.

While not designed as an acquisition vehicle, the program has, in two instances, led to investment or acquisition when the strategic fit aligned. Salyards emphasized that integration into Deere equipment is not the default outcome.

“This is one vehicle among many that we use to understand what innovative companies are out there,” he said.

“Ultimately, it helps us determine how well those technologies could fit for our ag and construction customers.”

Artificial intelligence, automation and data systems dominated nearly every session, from crop protection to robotics to biotech discovery.

However, beneath all that, one quieter theme kept surfacing.

A lot of the early, practical value of these systems is not in running machines. It is in interpreting data and turning it into recommendations.

Why it Matters: As AI tools take on more of the data work, farmers will still need trusted advice to turn those recommendations into decisions that work in their fields.

In fact, based on the discussions at the summit, that part of the conversation was in the rear view mirror. Much of the focus now is on what comes next — building systems that can act on those recommendations.

Soil tests, weather stations, satellite imagery, equipment data is familiar ground for agronomy. What is changing is how quickly and how consistently that information can be processed.

In one session about biotech discovery, speakers described AI systems that can design and refine experiments with minimal human input.

It is a long way from a Prairie field, but it is easy to imagine that same approach being used to improve plot trials or even guide on-farm decisions aimed at maximizing yield.

And it is already happening.

On the farm, that same capability is showing up in decision support — not perfect, not complete, but improving. These tools are getting better at taking large volumes of information and turning it into clear, actionable decisions.

From interpretation to action

And that raises a fair question. If more of that interpretation work can be done by a system, where does that leave farm agronomists?

The answer is not that they disappear. It is that the job shifts.

Research agronomists are not really in the crosshairs here. They are still building the knowledge base. The question is what happens to the people turning that knowledge into decisions on the farm.

That kind of agronomy has never just been about reading numbers off a report. It is about context: knowing the field, the farmer, the equipment and the risks they are willing to take.

A recommendation generated from data still has to be weighed against reality. Is the field fit? Does the timing work? What happens if the weather turns? Does it fit the rest of the rotation?

Those are not problems that go away with better models. In some ways, they become more important because more recommendations are coming, faster and with more confidence behind them.

What these systems may change is how agronomists spend their time. Less time pulling data together. Less time building base recommendations from scratch. More time stress-testing those recommendations, adapting them to local conditions and helping farmers decide what to act on and what to ignore.

There is also a practical layer to this that did not get as much attention on stage. These tools do not just appear on farms fully formed. They need to be set up, calibrated and understood. Someone has to translate them from a product into something that actually works in a field.

One discussion on soil health touched on a more basic issue: even something as fundamental as soil testing is not fully standardized. Results can vary depending on how samples are taken, handled and processed.

That is an opportunity.

It suggests there is still a role for the local private agronomist — someone who knows the region and their customers, understands local soil conditions, along with insect and disease pressure, and someone who farmers know personally and trust.

The role doesn’t disappear, it changes

It is easy to frame new technology as a threat to existing roles, but agriculture has a way of absorbing new tools and reshaping the jobs around them.

GPS did not eliminate the need for farm agronomists. Variable rate did not either. They changed the conversation.

This one feels different. These systems are starting to take on the interpretation work that has traditionally defined farm agronomy. However, the pattern is familiar.

The technology is moving quickly, that much is clear. However, it is still being tested against the same reality. Fields, weather and economics have a way of exposing weak ideas.

On-farm agronomy does not sit outside that process. It is part of it.

If anything, the need for people who can bridge the gap between what a system suggests and what actually works on the ground will only grow.

Until recently, moving a digital prescription from Bayer’s FieldView platform to a John Deere display typically meant downloading files to a USB stick and physically transferring them to the machine. Bayer and John Deere say a new integration between FieldView and John Deere Operations Center eliminates that manual step.

Why it Matters: Direct integration between agronomy software and equipment displays could simplify field execution and reduce setup errors.

The companies announced the upgrade at Commodity Classic in San Antonio last month.

With the new workflow, users create scripts in FieldView, select the relevant files and click “Export to Work Plans.” The jobs then appear in Work Planner within John Deere Operations Center, ready to run on connected equipment — all delivered remotely from one platform to the other.

For Prairie grain producers using both platforms, that means fewer steps between agronomic planning and field execution.

In an emailed statement, Bayer Crop Science said the goal is to simplify job execution and monitoring for customers working across both platforms, while eliminating the need for thumb drives and other manual steps that slow down field activities.

Chris Winkler, vice-president of digital software and solutions at John Deere, said the integration responds directly to customer feedback.

“Our mutual customers want streamlined workflows, not extra steps in the cab,” he said.

In practical terms, that affects both set-up time and accuracy. Removing manual file transfers reduces the chance of loading the wrong prescription, misnaming files or configuring monitors incorrectly — issues that can affect variable-rate seeding, fertility or crop protection passes.

The integration also changes how data flows back to the office.

Once jobs are completed, as-applied information moves back through Operations Center and into FieldView, allowing farmers and advisers to evaluate performance and adjust future prescriptions.

The companies say the capability is the result of feedback from farmers and advisers who wanted platforms to work together more seamlessly.

For Prairie growers, where variable-rate seeding and fertility programs are increasingly common in crops such as canola, wheat and corn, tighter integration between agronomy software and equipment displays could mean less time managing files and more time focusing on in-field decisions.

The capability is currently being introduced to select U.S. customers, with broader availability expected in the coming months. The companies have not yet outlined a timeline for Canadian rollout.

It’s that sort of cutting edge vision that saw Albrecht and Weed-It take home the 2026 Western Canadian Crop Production Show Innovation Award handed out at this year’s Crop Production Show in Saskatoon.

• Follow all our coverage of the 2026 Crop Production Show here

“This blows me away,” Albrecht said as he stepped of the stage, holding his crystalline trophy and its accompanying orange banner tightly.

“I’ve been doing this almost a decade now, so it’s really nice to see what I’ve helped bring to Canada. Weed-It is a mainstay in agriculture now, North America-wide.”

In addition to first place Weed-It, there are two runners-up:

Ground Truth Agriculture Benchtop MV/NIRS

This benchtop grain grading system uses machine vision and near-infrared spectroscopy to analyze Canadian Western Red Spring wheat. CWRS was tackled first because it’s notoriously one of the most challenging crops to analyze, according to company chief executive officer Kyle Folk.

The goal is to help producers receive accurate and consistent analysis of their crops.

The Benchtop system is already working on soybeans, red lentils, and amber durum, and future capabilities include corn, canola and oats. On-combine capabilities are also in the works.

Ultraview Technologies Inc. AgVision HD monitor

Company co-founder Brad Reykdal described improvements in image quality, ease of installation, future scalability and artificial intelligence compatibility as a few of the benefits offered by AgVision HD.

Reykdal says the AgVision HD system allows producers to watch four cameras live, with the ability to have eight connected in total.

Show organizers say the Western Canadian Crop Production Show Innovation Awards exist to showcase the products or inventions of exhibitors that drive innovation in the agriculture industry.

Why it Matters: GPS signals have become pervasive in agriculture, guiding all types of machinery and providing tracking of machinery and livestock. They also are used across the transportation and processing sector.

Iridium, which operates a mesh of low-earth orbit satellites and provides an alternative to GPS, now has a way to integrate its technology directly into GPS-enabled systems.

Iridium’s PNT system has required a box, which meant it was usable on ships and planes and for critical infrastructure but not in personal communication or the many other portable uses of GPS, including the type of units in tractors, sprayers and combines.

The company recently announced the Iridium ASIC, or application-specific integrated circuit. Think of it like adding a power take-off driven hydraulic pump to a planter to get enough extra hydraulic pressure to run more functions. It’s a dedicated, single-use upgrade.

Iridium hopes its chips will be installed across the economy, adding a second layer of security for GPS-enabled activities now that its Positioning, Navigation and Timing (PNT) signals can be managed by an eight-by-eight millimetre chip.

Iridium has a competitive advantage in supplying an alternative to GPS because its satellite network is a low-earth mesh network that covers the entire planet. Satellites that broadcast GPS orbit 18,000 kilometres or more higher than low-earth orbit satellites.

That means the well-meaning signal sent out to geoposition a ship, car or location tag in luggage comes from 20,000 km up, and the signal is weak when it reaches the Earth.

Being closer to the Earth means the signal that Iridium can send is 1,000 times stronger than GPS. It also means the PNT signal can penetrate buildings where there are challenges with weaker GPS signals.

“We all take GPS and other (global navigations satellite systems) for granted,” Iridium chief executive officer Matt Desch said during a news conference.

GPS are integrated into transportation, agriculture and cellular phones and is used to track everything from suitcases to cars. GPS signals are old, simple technology, not built for today’s more complex systems, and that has invited in bad actors.

“They were never designed to be resilient, and yet we built an entire digital economy on top of them. It’s like building a skyscraper on a sandcastle,” says Rohit Bragg, Iridium’s vice-president of commercial PNT.

What’s the concern?

There have been several recent examples of GPS or GNSS signals being blocked or hijacked.

Shipping managed by Qatar was shut down for 48 hours due to GPS jamming, says Desch.

On Aug. 31, a plane carrying Ursula von der Leyen, president of the European Commission, had its GPS signals blocked as it approached the runway in Bulgaria, and pilots had to rely on other group-based signals to land the plane.

The commission blames Russian interference.

About 1,500 flights per day experience some GPS spoofing, says Bragg, citing a report by OpsGroup, an aviation safety organization.

It’s not just GPS blocking that’s an issue, but also deception. Bragg gave the example of Santo Spirits, an American alcohol company, which thought trucks were on the correct route according to the GPS signals they were reading, but they never arrived, and in fact had long been off the expected route.

This example created concern throughout transportation systems, and created opportunity for new solutions, such as that created by Iridium.

A new mobile app in development at the University of Saskatchewan promises to identify field insects instantly, show local populations on a live map, and deliver management advice based on crop, region and weather.

Why it Matters: Tracking pests and beneficial insects in real time could help farmers make quicker, better-informed pest management decisions and cut unnecessary pesticide use.

The app, called IPPM Now, is expected to combine artificial intelligence (AI), geospatial data and entomology expertise to turn a smartphone photo into real-time agronomic insight. Its developers say it recognizes both harmful and beneficial insects, from flea beetles and grasshoppers to pollinators and lady beetles, with more than 90 per cent accuracy.

Farmers could therefore learn not only what insect they’re dealing with, but whether pressure has reached economic thresholds and what conservation steps might protect beneficial species.

For developers, the goal is to pull together information that has long been scattered across research programs, scouting reports and grower experience.

“It will be super useful for farmers, agronomists and scientists scouting insect pests,” says project lead Teresa Aguiar, a PhD student at the University of Saskatchewan. “Scouting takes a lot of time, and the information from researchers, agronomists and farmers is often disconnected.”

From photo to field map

Each image submitted through the app is tagged to a rural municipality, not to an exact GPS point, to protect user privacy. Those records build a colour-coded map that shows where pests, pollinators and biocontrol insects are active. The development team plans to integrate local weather data so future versions can forecast outbreaks and pollinator activity.

“We want a practical tool that integrates insect identification, spatial reference, data collection and management recommendations in one platform to make decisions with all the variables involved in pest management,” Aguiar says.

Smart traps and sweep-net scouting

Alongside the app, Insect Track Solutions, the Saskatoon-based startup commercializing the project is testing a smart trap that marries a sticky card with a small camera. Set in a field, the trap automatically photographs insects and uploads images to the same AI model used by the app, identifying and counting adult insects without anyone having to check the card manually.

Because sticky cards only capture flying adults, Aguiar’s team also designed a simple, low-tech workaround for ground or juvenile stages.

“To solve this problem, make a sweep and then put that sweep in a ziplock bag with a white background and take a picture of that ziplock bag using our mobile app,” she says.

The model can then identify and count nymphs and instars in the sample, giving a fuller picture of population levels.

What it can do today

The prototype version already supports canola and wheat and recognizes 10 key insect groups common to the Prairies. Pests include lygus, flea beetles, grasshoppers and weevils, while beneficials include lacewings, lady beetles, hoverflies, bumblebees and honeybees.

The app draws on field data and photo libraries, including data and images supplied by Manitoba entomologist John Gavloski, to keep improving its accuracy toward species-level ID.

Future updates will broaden crop coverage and add weather and growth-stage links to help predict pest risk or pollinator timing.

Beta testers wanted

Before IPPM Now officially launches next spring, the developers are inviting farmers, agronomists and crop scouts to test the app this winter. Beta users will get early access, provide feedback on design and function, and can volunteer to host free smart-trap trials in 2025.

Aguiar says user input will guide the final version.

“We are sending beta testing invitations. If you’re interested, we can put the app in your phone, early access to give us feedback and help us to shape the app and get it ready for next season.”

Farmers and agronomists interested in early testing or hosting field validation sites can contact Insect Track Solutions Inc. via email.

There’s been growing interest in precision 4R fertilization over the past few years, said Mario Tenuta during the Canola AgriScience Cluster Research Roundup webinar earlier this summer.

The 4R movement has been the key framework for Fertilizer Canada, received official support from Manitoba Agriculture and been touted and incentivized by various provincial and federal commodity groups, funding programs and organizations such as Farm Credit Canada.

But Western Canada and even individual farms have a lot of variation in their soils and topography, Tenuta said, and that’s a challenge for farmers looking to adopt 4R.

Tenuta has been at the leading edge of multi-stakeholder research into 4R nutrient management.

Possibly the biggest takeaway from one study has been consistently higher nitrous oxide (N2O) greenhouse gas emissions coming from the lower areas of fields.

“(Fields) have underlying landscapes, and that affects the return on investment of nitrogen fertilizers and changes nitrogen losses (and) response to fertilizer application, depending where you are on the landscape,” said Tenuta.

The hypothesis of the study, he said, is that by saving nitrogen and not losing it, farmers don’t have to apply as much nitrogen fertilizer. That can save producers cash while minimizing greenhouse gas emissions.

”Nitrous oxide is emitted more so in depression areas because it’s wetter. And so we know that certain areas of the landscape have greater potential for N2O release than others,” he said.

“So when we’re managing a strategy to reduce emissions — and where some 4R practices cost money and potentially even actually spoil some profits — it’s important to know (if we can) tailor the 4R practices to those regions or areas of the field which give us the most return in terms of greenhouse gas reductions and also less impact on profitability.”

The study — which has two more years to go — is a collaboration among the University of Manitoba, University of Saskatchewan, the Indian Head Agricultural Research Foundation and the three provincial Prairie canola grower associations under the umbrella of the Canola Council of Canada.

Tenuta’s co-investigators included Philip Harder, Rich Farrell, Tristan Skolrud and Chris Holzapfel.

“The research is done on commercial farm fields and … the agronomics, treatment application and so forth is actually done by farmers,” said Tenuta.

Three research sites were chosen — two in Saskatchewan (Saskatoon and Indian Head) and one in Manitoba (Brandon).

Each field was divided into 10 management zones based on soil properties such as soil nutrient levels, water regime, water movement, drainage and topography. This was verified and fine tuned by ground-truthing.

Then came the development of SWAT (soil, water and topography) maps by Saskatoon-based ag tech firm Croptimistic Technology. These provided variable rate fertilizer prescriptions based on the management zones and their expected yield.

“(We) take some samples — lots of samples actually — for nutrient analysis, find out what’s present in the soil before planting and do a yield analysis in terms of seeing what has been the previous responses to nitrogen fertilizer in these different management zones. And they use all that to fine-tune a prescription map,” Tenuta said.

Then came different fertilizer treatments and the measurement of resulting emissions.

Zone treatments included a flat rate of urea based on a general soil test, variable rate nitrogen based on the SWAT prescription map, a variable rate urea and SuperU mix, plus that same treatment minus 10 per cent of the nitrogen.

The variable rate urea and SuperU mix were considered the precision 4R treatment.

The first tested location was the commercial field near Brandon in 2023. N2O emissions were found to be highest in the depression zones. However, it was a dry summer, which likely weighed on emission levels, said Tenuta.

The trial at Indian Head in 2024 saw a large drop in emissions under variable rate urea in the depression zones. The mid-slope zones saw a benefit from the precision rate and precision rate with reduced nitrogen.

Canola yields in the knoll tops increased with a SWAT map prescription of nitrogen and with precision 4R treatment. Yield in the lower zones was generally consistent and at par with some of the midlands under that same treatment.

The Saskatoon test saw N2O reductions in the mid-slope zones under variable rate urea.

“When we see this reduction with the precision practices, that’s quite a reduction, actually, in terms of N2O emissions,” said Tenuta.

Warren Ward, an agronomist with the canola council, asked Tenuta about the capacity of producers to invest in precision 4R equipment.

“One of the challenges I always hear about for uptake on new practices is logistics and equipment in addition to cost. So I’m just wondering, in your opinion, what’s the feasibility this precision 4R approach at this point in time?” Ward said.

Tenuta said the cost hurdle is becoming smaller as time goes on, especially as precision tech continues to be packaged with farm equipment.

However, there’s a bit of a class divide. He said producers with “big acres” and economies of scale are not only the ones most likely to afford this decked-out equipment, but benefit from it the most. Interest is also a factor, particularly among younger producers.

“It’s fun for them,” Tenuta said.

“It’s going to be those folks that are going to be the ones that this technology appeals to, and they’ll have the equipment.

“The ones that don’t have the equipment that are dealing with seeders and drills and things like that from the 1990s … they’re going to retire that way and they’re not going to be doing this.”

The project funders include Alberta Canola, SaskCanola, Manitoba Canola Growers and the Sustainable Canadian Agricultural Partnership through the Canola Council of Canada Science Cluster.

They were robots, used artificial intelligence and killed weeds with laser beams.

Field-applied plant cell exploding technology was one of the cutting edge agricultural ideas presented at the Bradford AgRobotics demonstration day earlier this summer.

Tech entrepreneur David Tao, chief executive officer and co-founder of Ontario-based BH Frontier Solutions, presented a compact BHF ElectricWeeder, while Dutch ag-tech firm Pixelfarming demonstrated a modified laser model.

Haggerty AgRobotics president Chuck Baresich said the Pixelfarming Laser One-i prototype, adapted for the Orio NAiO’s frame, arrived in March. NAiO is an autonomous field robot in its third year at the Ontario Crops Research Centre in Bradford.

It features six laser units that use high voltage to energize a carbon dioxide tube, creating a focused beam that boils the liquid inside plant cells, effectively destroying the plant.

“The lasers will shoot forward and back,” Baresich said. “It will determine the optimum angle as you’re moving forward to hit the various weeds.”

Rydan Chalmers, a Haggerty AgRobotics Systems Engineering intern, used 35 images of multicoloured confetti to demonstrate how easy it is to program the interface’s crop module to target specific plants or, in this case, confetti colours for eradication.

Baresich clarified that plant images at various growth stages and under different lighting conditions are necessary to develop a comprehensive crop model, noting that the rapid pace of technological advancement and the availability of open-source tools are accelerating the process.

“You can load five (models) at a time,” Chalmers said.

“We have a few examples we’ve been testing in the field. And you want to make sure you have the right laser selected, (and) how much energy is going to each laser.”

For the demonstration, Chalmers set the lasers to 80 per cent, which, while producing a satisfying flame-out, was excessively high because most weeds and grasses are effectively terminated at 25 to 40 per cent power.

Baresich warned that there is a low chance of starting a small fire if a field is dry or of damaging a neighbouring crop if the heat isn’t properly adjusted.

The LaserOne-i’s speed and effectiveness are optimized for dime-sized plants or early emergence because larger plants slow the machine significantly.

“When you have a very, very tiny weed, it takes only milliseconds of heat to cook that,” Baresich said.

“That’s how you get the speed of the machine up.”

Field trials comparing 90,000 weeds to a million weeds per acre showed a significant slowing of the machine due to the number of shots per second per acre, a common issue when dealing with grass because it’s prolific, Baresich said.

Depending on the heat level and dwell time required to eliminate a plant, the three lithium phosphate batteries provide a four-hour run time, with the ability to add additional batteries.

BHF ElectricWeeder is an autonomous robot that selectively kills weeds with high-voltage electricity. It is guided by artificial intelligence using an integrated multi-camera computer algorithm that differentiates between weeds and crops.

“The algorithms are trained to recognize the crop; anything different from that would be considered a weed,” Tao said, adding that the crop model trains with 100 images, while shape and morphology changes require additional training.

The robot collects data automatically, improving the AI’s accuracy and increasing its speed without the need for farmer oversight outside of the initial RTK to plot out the rows.

“Once it detects the weeds, it will guide the robot arms underneath and basically target the weed and zap it,” said Tao, adding it’s part of an onion crop trial at the research centre.

During the demo, the robotic arms moved swiftly and methodically through the crop, unleashing an electric current that often produced a visible puff of vapour as the weed’s cells burst.

The key difference between the ElectricWeeder and other robotic options is its ability to dispatch weeds at any size and growth stage.

“It has two modes of operation tailored towards both low weed density fields and high weed density fields,” Tao said.

“It will maintain the same efficiency at killing weeds.”

The 1.2-tonne machine can weed up to 10 acres an hour and operate in most weather conditions 24/7, but like traditional iron, mud can be problematic.

“A bit of rain, a bit of drizzle, a bit of moisture is fine, but not too much,” he said.

“Not like a muddy kind of pouring rain where you have ponds in the soil.”

The ElectricWeeder comes in autonomous, tractor-pulled and tractor-liftable models, with coverage from five to 20 feet and swath widths of 60, 120 and 240 inches, plus an adjustable row spacing of 60 to 88 inches. The sub-millimetre precision and 99 per cent weed kill rate make it more efficient than a 100 person hand crew.

Tao said the US$150,000 commercial versions will ship near the end of the year.

Robot weed control will never be a one-time pass, said Baresich.

“Part of the economics of the machine is how many times do I have to go through that same field, and can I get it done?” he said.

“In comparison to hand-weeding crews, where the costs are $500 to $1,000 an acre, these are very economical to run.”

He said chemical weed treatment remains cheaper; but in cases of herbicide resistance and hand pulling, the return on investment of laser or electric weeders pencils out.

“I think the reality is that even though the ROI might not work, we’re going to be forced into the robotic technology just because of weed resistance,” said Baresich.