Industry collaboration key to 'cutting-edge’ medicinal cannabis research

A PhD research project through the QUT Centre for Robotics is developing an automated system to prune medicinal cannabis plants - something that could be a game-changer for an industry that struggles to find skilled workers. 

QUT Senior Lecturer Dr Chris Lehnert is supervising the project which is using computer vision, artificial intelligence (AI) and robotics. 
“Leaf pruning is labour intensive, has to be done throughout the growing cycle and is very costly,” he said.  

“Automation has the potential to reduce costs, but we believe there’s other value in terms of monitoring the crop while pruning so you can get some feedback on growths, or even the yield.”  
Medicinal use of cannabis was legalised in Australia in 2016 and while it’s highly regulated the industry has been growing rapidly. IBISWorld analysis shows the sector’s revenue has more than doubled in the past two years to $350 million in 2023/24. 

In 2023 Australia produced more than 26.5 tonnes of medicinal cannabis, with two tonnes of that exported, while more than 42 tonnes was imported into the country. 
Mechanical and automated pruning is available for other horticulture crops – think saws cutting straight lines through grapevines. But when it comes to leaf pruning medicinal cannabis, precision is key, Chris explains.  

“You have to be very precise because the buds or the flowers are very near, close to the leaves,” he said. 

“There's a been a huge shift in AI and that means you can use vision systems or camera systems to understand exactly what you're seeing and, through the use of algorithms, inform a good decision about where to cut.” 

Along with the system to identify where to prune, the research is designing new tools to do the job, with hopes of patenting the technology and commercialising in the future. 

RELATED ARTICLE: How are robots used in agriculture? 

When it came to field testing the technology and algorithms on real plants, the team turned to AgriFutures grow^AG. to find industry partners.

 The online platform has listed more than 3000 research projects, with 200,000 global users, and has facilitated more than 1,500 connections.  

QUT Senior Manager, Industry Engagement (Agricultural Sciences and Bioeconomy), Melissa Nugent said grow^AG. put them in touch with the right people at the right time. 

“You're not scrambling around talking to people who aren’t interested at this time, you're actually talking to people who are looking for you, it just cuts out all the noise,” she said.  

“We had no contacts in medicinal cannabis at all. Using grow^AG. we went from cold, to three prospective partners who wanted to talk to us, to finding the right fit in a matter of two months.” 
Mel said the support from the grow^AG. Concierge Service made listing on the platform easy. 

“The way the team facilitates pulling together all the key information for that Expression of Interest makes it simple and effective,” she said.  

“It's super quick compared to prospecting, going to conferences, scouting sites, trying to be where those industry people might be, building those connections. It just cut through all of that.” 

RELATED ARTICLE: Navigating and finding your tribe in the agrifood tech and innovation ecosystem 

Chris was impressed by how quick and easy it was to engage with growers through the platform. 

“Working in agricultural research you want to get in contact with growers, but they don't have much time,” he said.  

“The value of the platform is that it gives them the ability to easily let us know that they're interested in talking, instead of us trying to reach out to them and it not being the right fit.” 
The connections forged through the platform are delivering results. 

“The immediate value is having a conversation with the cannabis producers to understand the industry problems and the diversity in the way the industry is working,” Chris said.  

“It was great to understand their individual problems and work through with us at the early stages.” 

He said finding an industry partner has been crucial to the research because it’s meant that rather than trialing the technology on a model or crops similar to cannabis, the student has been able to test the robotic system on the real thing. 

“The partner has been happy with the initial tests and has been providing feedback,” Chris said.  

“The next stage that we’re working through is potentially how we could arrange the plants to assist in robotics and the grower seems to be very willing to adapt or change, if the robotics is commercially available in the future.” 

Mel added that the engagement through the grow^AG. platform has been useful in showcasing what’s possible when it comes to robotics and automation. 
“In our discussions with those partners, they weren't familiar with the most recent technology developments and the opportunities that automation could present them as a specific business,” she said.  

“What are the tasks versus where the research is up to, and for the more complex tasks, what is on the horizon for automation?” 

More field trials are planned, and Chris believes there’s opportunities to take the technology further.  

“Towards the end of the PhD, we would hope to discuss what opportunities we could have to either commercialise or take it to the next level with the industry partner and they've seemed very interested in doing that.” 

When that time comes, grow^AG. would be something we could advertise on, not just for industry partners in cannabis, but technology partners as well,” Chris said.