Marijuana growers are investing heavily in automation technology to remain competitive and boost efficiency, but pitfalls abound
by Omar Sacirbey
There’s a war on the horizon for cannabis cultivators.
As the marijuana industry matures and evolves, wholesale cannabis prices are plummeting in some markets – especially those that allow dozens or hundreds of cultivation operations. The result: Growers in these states are facing new competitive and financial pressures. Profit margins are shriveling, cannabis is becoming a commodity and big players are starting to put the squeeze on their smaller counterparts by building economies of scale.
These trends are particularly acute in the recreational market, but they’re also playing out in medical cannabis states as well. And they’re expected to accelerate in the coming years as the industry marches forward. In the future, growers that don’t or can’t adapt to these changes will get gobbled up or destroyed completely by the competition.
With this as the backdrop, cultivation companies increasingly are turning to automation to cut costs, gain an edge over the competition, bolster control over the grow process and improve efficiencies.
Mechanization can accomplish many of the things that people now do manually, such as turning up the thermostat or watering and feeding the plants. If smartly implemented, automation can save businesses tens of thousands of dollars over the long haul and improve the quality of the plants.
But diving headfirst into automation blindly can be a big mistake. Growers must be selective about which processes they choose to automate, and they must remember to crunch the numbers, particularly when it comes to production costs. While automation can be a great solution for many situations, growers have been known to spend six-figure sums on technology they ultimately didn’t need – or that failed them. Trial and error is sometimes a given.
Shane McKee, chief cultivator with Shango Premium Cannabis, a marijuana brand and dispensary chain in Portland, Oregon, noted that automating the irrigation system in a large grow can be especially tricky.
McKee spent hundreds of thousands of dollars on failed or expired automation equipment in general, including irrigation. “There have been things that we bought that didn’t work, and things that we bought that we’re still using,” he said. “So there’s been a lot of research and development along the way – things that we spent money on that we’re no longer using.”
Cultivators have different opinions on what should and shouldn’t be automated and how different technologies should be implemented. What they do have in common is how they arrive at their decisions: They review their costs and the problems they’re trying to solve, and they determine how mechanization can help. Growers that take this approach to automation are much more likely to succeed.
Autumn Rose Karcey, president of Cultivo, a cannabis cultivation consulting firm in Los Angeles that specializes in automation, predicts that growers may not have a choice when it comes to employing technology, especially those that operate a big grow.
“In five years, it’s not going to be a choice of ‘do I automate or not,’” she said. “In a large-scale operation, automation is key for survival. The Nabisco cookie factory isn’t going to question whether they use automated equipment. You just reach a point of growth where you have to automate in any manufacturing practice, in any industry.”
Even small growers can benefit from mechanization, Karcey said. If a fan malfunctions when you’re out of town but you have a sensor that can alert you, you can call someone to fix the problem.
Moreover, the necessity of automation won’t be driven just by size, but profit margins. The 30% wholesale margins many growers have enjoyed won’t last: They are already dissipating in mature markets like Colorado and Washington state. When those margins decline, which players survive will largely depend on who cuts costs.
“Automation is going to play a key role,” Karcey said. “If it costs me $500 to produce one pound of cannabis and my neighbor $900 to produce one pound – based on the fact that I have automated equipment – who’s going to survive?”
Automation for Your Environment
What to automate first? Most experts say you should aim for the grow’s environment – air conditioning, heating, dehumidifying, humidifying, watering and feeding.
“The most important thing is environmental controls with multiple redundancies,” said McKee of Shango, which has a total of about 200,000 square feet of cultivation space at its sites in Oregon, Washington and Nevada. The company produces Chong’s Choice flower, Shango-branded edibles and other products.
If you’re going to automate your facility, you also may as well invest in software that monitors temperature, humidity, CO2, irrigation and other factors at your site. The technology alerts you via text or email when any of those factors get too low or too high.
For example, McKee was alerted after one of his Portland sites lost power. He went there to make sure the backup generators and controls had kicked in and were functioning until the main power returned.
He said it’s “critical” to have backup generators to ensure your systems continue operating in the event of a power outage.
It’s worth remembering that while automation software is powerful, it would be risky to use by itself without backups. There are many ways to implement such backups.
“There’s no one system that you can trust when you have 500 pounds of product in one room,” McKee said. “You build in redundancies. If one thing fails, you can catch it another way.”
At Shango, for example, McKee has a primary set of digital sensors to monitor the environmental conditions in his sites and regulate equipment. If those sensors fail, he has backup analog sensors that come on and regulate the environment as need be.
There are other ways to minimize risk. If you have a room that requires 25 tons of air conditioning, for example, you don’t use one 25-ton air conditioner. Instead, you try five, 5-ton units. That way, if one air conditioner goes down, the others will suffice until the broken unit is running again.
“If we’re down 20% for a few days, it’s not going to kill me,” McKee said, adding that the crop isn’t damaged.
Compared to climate and humidity, watering and feeding are more difficult to automate, McKee said. But it can be done. Like climate monitors, water and nutrient monitors allow growers to watch and adjust how much of both their plants are receiving.
But be aware that automating these may require some trial and error.
“The toughest struggle we’ve had is automating the plumbing on large-scale grows,” McKee said. “When you get into these large facilities where the water and nutrient sources are far away and the rows are very long, that is where we’ve gone through four or five different types of systems that didn’t work.”
At one site, for example, plants on one side of an aisle were receiving half the amount of water and nutrients that the plants on the other side were getting. The reason? The plants on the well-fed side were closer to the nutrient reservoir than the plants on the underfed side, so that the latter received food more slowly. The solution was to extend the hoses on the well-fed side of the aisle so water and nutrients traveled an equal distance to plants on both sides.
You can also receive notifications about whether you’re supplying too much or not enough individual nutrients. After receiving an alert, for example, you can give your plants a small dose of water until you correct the nutrient issues.
The size of your grow and the plumbing will play a major role in how you automate your watering and feeding.
“It really comes down to the room sizes and how central the nutrient reservoir is, whether you built from the ground up and you were able to lay the plumbing underground, or whether you have the plumbing overhead because you’re improving an existing building,” McKee said.
Trimming machines are another example where cultivators have found relief through mechanization.
Richard Abromeit, CEO of Montana Advanced Caregivers in Billings, decided to buy a trimming machine for about $8,000 in 2014 after becoming fed up with the high employee turnover for the trimmer position.
The company started testing the machine and adjusting its controls, then advertised the change on social media so patients would know about the cost savings.
“Trimming was revolutionary for us,” Abromeit said. “The machine does the work of five people, and we can pass the savings on to patients.”
Abromeit did have concerns that machine trimming would hurt the product. But those fears proved wrong.
“Quality hasn’t suffered,” said Abromeit, whose employees do the trimming while the plants are still wet with oil and moisture. That way the trimmer won’t beat the trichromes off the plant.
The only added cost Abromeit has put into the trimming machine is $750 for a new fan motor. “Had we cleaned it right,” he said, “I think that piece of machinery would have worked for us another five years easy.”
Now his team cleans the trimmer about every 10 days. It takes one person two hours to clean it. Ideally, you want to do it at the end of the shift when you can let the parts sit overnight with the cleaner on it and then rinse it off in the morning with water. Then you put it back together in about 10 minutes with T-wrenches.
It takes about two weeks for employees to learn how to use the trimmer, which can’t be left running unattended. “There’s a learning curve and you notice that your trimming gets better as you use it,” Abromeit said.
The cannabis is fed into the hopper and then flows into the trammel and into the trimmer. Trimmer machine operators need to know that separate strains must be fed into the machine at different speeds. That’s because various strains have differing characteristics – their density, how leafy they are and how stringy they are. All of these influence how the strains go through the trimmer.
When Not to Automate
While growers who have tried automation generally endorse it, many also believe that not every process requires automation. Moreover, many growers disagree over what should and shouldn’t be automated.
For example, while Abromeit swears by his trimmer, he doesn’t leave watering and feeding to automation because he likes to give different strains varying amounts of nutrients. That can be harder to regulate in an automated system. So he does it by hand.
“One plant strain will take nutrients at a different rate than another – say a Kush as opposed to a faster-growing sativa like a Super Silver Haze, which needs to be pounded with nutrition,” Abromeit said.
The only way to achieve selective feeding without hand-watering is to have multiple nutrient reservoirs marked for specific strains. But reservoirs take up a lot of space, and having more than one may not be a good option when you have a craft grow business like Abromeit’s facility.
“Because of these differences, we do it by hand,” Abromeit said.
Cultivo’s Karcey, by contrast, considers hand-watering a waste of time and is skeptical about machine trimming.
“There are some things that you can automate, and there are some things you can never duplicate, like the hand trim,” Karcey said. “If I put everything through trim machines, I wouldn’t be competitive in a market like California. You have to have hands on that.”
McKee feels that while his backup generator is critical in supporting his climate control equipment, he doesn’t use it as a backup for lighting.
“We’re not worried if they don’t get light for eight or 10 or even 12 hours,” he said. “We’ve got a few backup lights in there. We’ve got some heat in there, dehumidification, just the absolute minimums that would be necessary to keep the plants in an OK state.”
Deciding on Automation
If different growers offer opposing advice on when to use automation, how should a grower who is new to automation decide what to do?
You start with numbers, especially production costs, Karcey said. You must know your cost per pound of production, how much electricity your building uses, your air conditioners’ BTUs and the cost of your water, nutrients and labor, among other considerations.
“You have to ask, ‘What does it take to grow this cannabis?’” Karcey said.
Another good reason to know your numbers? Engineers will sometimes recommend the equipment and setup that makes them the most money, and not necessarily what is best suited for your site. Knowing the numbers will help you make the appropriate choice.
“You really have to ask yourself, ‘Is this what I need?’” Karcey said.
Another way to decide is by experimenting. If you’re considering introducing automation, start in one room. That way you can minimize the impact of mistakes. And, if it works, you can scale up.
Cultivators with the space may even consider sectioning off a part of their buildings and reserving that area for experimentation, Karcey suggested.
Another way to decide whether a certain situation merits automation: Does it relieve you of unskilled labor chores and give you time to spend in your garden examining your plants?
“When a guy gets caught up doing chores and his job becomes doing chores, he pays attention less,” McKee said. “If you don’t have all that labor to do, you’re able to pay attention to details.”
Karcey agreed. “The idea of automation is not to give growers a vacation,” she said, “but it’s so they can do their jobs more effectively as cultivators. It’s to give them more time to spend with the plant.”