Before You Bet Production on One Machine

If one machine carries 100% of your production plan, you do not simply have 25,000 units of daily capacity.

You have 25,000 units of capacity on the condition that one machine keeps running.

That is a risk every pre-roll producer should understand before approving a major equipment purchase.

The machine will eventually need cleaning, maintenance, adjustment or repair. That does not make it a bad machine. The real question is what happens to the business while it is not producing.

The machine is down. Payroll is not.

The flower may already be milled. The production crew is standing there. Packaging is waiting for finished pre-rolls that are not coming. A retailer promotion is scheduled for Friday, and somebody is on the phone trying to determine whether the problem needs a reset, a replacement part or a technician who is not in the building.

That is what downtime actually feels like.

It is rarely one quiet machine and one neat line on a maintenance report. It moves through the facility. Material waits. People get reassigned or paid to stand by. The next SKU gets pushed back. Overtime appears on Saturday. A co-packing customer starts asking whether the order is still shipping.

Even after the machine starts again, the production day does not magically return to normal. The team may need to reload material, recalibrate the settings, run test units, clear QA and then work through whatever piled up while the line was stopped.

The repair time and the business recovery time are not always the same thing.

That is the warning behind Part 9:

Never model only your production capacity when everything is running. Model what the facility can still produce when the main unit makes zero.

We know this story because we lived versions of it before building FlexPak, and we hear it regularly from operators who already own full automation.

In fact, one of the fastest-growing groups of FlexPak customers already owns a fully automated pre-roll machine.

That sounds a little backwards until you hear why they called.

Some need production backup for the days their automated line is being cleaned, adjusted or repaired. Others start with that plan, run the systems side by side and eventually move most or all of their production to a modular FlexPak setup.

In more than a few facilities, the backup stopped being the backup.

That pattern changed the way we look at an equipment comparison.

Output, labor and price all matter. But another number tells you how exposed the operation really is:

How many finished, accepted pre-rolls can the facility still produce when its main machine produces zero?

Call it remaining capacity. Call it survivable capacity. The name is not important. The number is.

If one machine carries the entire production plan, one machine going offline can remove the entire production plan. If the same volume is divided across two genuinely independent cells, one problem may remove half the capacity while the other half keeps moving. Spread it across three equal cells and roughly two-thirds remains when one stops.

Modularity does not make downtime disappear. It changes the size of the hole downtime leaves behind.

Suppose a facility normally produces 25,000 finished, accepted pre-rolls in a shift.

If all 25,000 depend on one machine and it is down for the shift, the floor produces zero against that plan. If the same capacity is split across two independent 12,500-unit cells and one is down, 12,500 units can still move forward.

Both facilities had the same capacity yesterday.

Only one still has meaningful capacity today.

This is standard manufacturing logic, not something we made up to sell pre-roll equipment.

Researchers at the University of Michigan have studied reconfigurable manufacturing systems for decades. Their work describes modularity and scalability as ways to add or subtract production resources when demand changes. It also says reconfigurable systems should be designed to respond to unexpected machine failures instead of letting one long failure define the entire factory.

NIST looks at equipment productivity through availability, performance and quality. Its manufacturing research lists failures of equipment, tools, sensors and control systems among the disruptions that damage performance. It also connects productivity directly to equipment availability and maintainability.

That is a different conversation from top speed.

It is about how the production system behaves on the day the top-speed number is unavailable.

The automated pre-roll manufacturers themselves are fairly open about that reality.

RollPros recommends a 30- to 45-minute cleaning after a Blackbird has run for about eight hours, plus a deeper 45- to 60-minute cleaning weekly or as regulations require. Its support page says every second the Blackbird is not operating costs the customer money, which is why the company provides remote diagnostics and keeps consumables ready to ship.

Hefestus makes a similar point in explaining its 24/7 support for the AuraX and AuraOne. The company says even reliable machines can develop issues that need technical help and calls downtime “the enemy.” It reports that most problems can be handled remotely, often quickly.

Canapa’s published JuanaRoll case study describes remote troubleshooting during commissioning and a customer learning which spare parts needed to be kept on hand so the system could remain operational.

Those are not admissions that the machines are poorly built. They are signs that the manufacturers understand commercial equipment. Cleaning is normal. Parts wear. Adjustments happen. Good support matters because good machines still have days when they are not producing.

Custom Cones USA has been especially direct. In a Ganjapreneur equipment guide, the company recommended multiple smaller machines before one large machine so production can continue when one needs cleaning or maintenance. Its current automated-equipment guide makes the same case for owning multiple automated machines.

That distinction matters. Modularity is not exclusive to semi-automation. A company can build a redundant floor with multiple fully automated lines too.

The question is whether the economics allow it.

Duplicating a six-figure production asset is a very different capital decision from adding another modular production cell. Plenty of companies can justify one fully automated line. Far fewer want to buy a second one mainly so the first is allowed to have a bad day.

That is where semi-automation can materially change the risk calculation.

Our observation about backup buyers comes from FlexPak’s own customer base and sales conversations, not a third-party market study. We are stating it that way on purpose. But it has happened often enough that backup capacity is no longer an unusual reason companies come to us.

Perpetual’s experience shows why this gets bigger than a maintenance headache.

Before FlexPak, CEO Dan P. had already worked around several complicated pre-roll systems. At a previous company, he dealt with an ActionPac system that cost more than $400,000. When it needed repairs, the process could involve specialized mechanical support, technician travel, replacement parts and extended downtime.

At Perpetual, the company later invested in a RollPros Blackbird. Dan and two team members traveled to Washington for training, expecting approximately 4,000 to 5,000 pre-rolls per day. Once the machine was installed, the learning curve, material sensitivity and finished quality became larger problems than the brochure had suggested.

Perpetual’s monthly pre-roll sales had previously reached roughly 120,000 to 140,000 units. During the Blackbird period, Dan reported that sales fell to approximately 35,000 to 40,000 per month. He estimated the decline at about $250,000 in monthly sales.

It would be too simple to blame every lost unit on mechanical downtime. Material preparation and inconsistent finished quality were major parts of that story too. But that is exactly the larger risk: when one production system struggles, every issue starts landing on the same calendar and there is no second dependable path to absorb the work.

Dan put the downtime cost in much plainer language:

“

“Two days of downtime hurts. I notice that in my end-of-month sales numbers.”

After installing one FlexPak Pro, Perpetual rebuilt the category to approximately 130,000 to 140,000 pre-rolls produced and sold per month. During one recorded week, two primary FlexPak operators produced 69,519 pre-rolls, supported by a material-prep employee and a separate tubing, labeling and case-packout team.

Dan has since made FlexPak part of Perpetual’s standard equipment plan for expansion into other states. Again, those are client-reported results and not an audited controlled comparison. What makes the story valuable here is the connection between production continuity and sales. The line does not have to be down for a month to damage a month.

Rebel Roots Farms shows the other side of modularity: installed capacity does not have to equal deployed capacity every month.

At peak, Dean S. reported that Rebel Roots built a modular assembly-line workflow around five FlexPak machines and produced as many as 2 million pre-rolls in a month. A sixth machine was available when orders required all six, or when one of the regular five needed to rotate out for cleaning or tuning.

When orders slowed, the team could drop down to two, three or four machines and bring the others back online as demand returned. They did not have to keep one oversized system running just to justify owning it.

That gave Rebel Roots both surge capacity and maintenance capacity. Five machines running with a sixth ready is a very different risk profile from one machine carrying the entire schedule. If one production cell came offline, output did not fall to zero.

Dean described how the operation scaled one unit at a time:

“

“Depending on the type of order and how much margin I had, it would be between four days and two weeks before I had the full ROI and could go get another machine.”

These are Rebel Roots’ client-reported historical results, not a universal production promise. The bigger lesson is the operating model: add capacity when the orders justify it, park capacity when they do not and keep a spare path available when maintenance shows up uninvited.

Surterra’s Florida operation, part of Parallel, gives us the clearest documented example.

Before standardizing on FlexPak, the facility had already invested in an ActionPac RollMaster 420 and an eight-channel JuanaRoll. David B., the operations manager we interviewed, reported that the JuanaRoll normally produced about 12,000 to 15,000 pre-rolls during a scheduled 10-hour day with five to six technicians involved. The ActionPac required eight technicians, including a large back-end crew to finish the cones.

They later built two separate FlexPak Max lines, with one machine in each line and four technicians assigned to each complete process. One line could handle 0.35-gram dogwalkers while the other ran 0.5-gram and one-gram cones.

David reported approximately 25,000 to 27,000 pre-rolls per line and 48,000 to 50,000 combined during a 10-hour production day, with less than 1% rejects. Those are client-reported results and have not been independently audited.

The output improvement is obviously important. The architecture may matter just as much.

The new setup did not put Surterra’s entire production plan inside one machine. It had two lines that could run different SKUs at the same time. A workflow adjustment on one did not require stopping the other. One line could be cleaned while the other kept working. If a component on one line needed attention, the second line still existed.

The ActionPac and JuanaRoll were eventually retired from regular production. FlexPak became the operating standard.

That story will not repeat exactly in every facility. Another operator may have excellent results with one integrated automated line and never want to change. Someone else may run several machines and discover that all of them still depend on one grinder, one air compressor or one closer.

That is the part people miss when they hear the word modular.

Two machines do not automatically create a resilient production system.

If both machines depend on the same grinder and the grinder fails, the filling equipment is now very expensive furniture. If every filled unit must go through one closer, the closer is the real single point of failure. The same problem can hide in a sifter, air supply, electrical circuit, software license, replacement part or the only technician who knows how the line works.

You have to map the entire order, not just the machine in the middle.

Start at material preparation and follow the product through filling, packing, closing, QA and whatever point your team calls finished. At every step, ask a plain question:

If this stops, what else stops with it?

That gives you the actual failure boundary.

Then calculate more than the repair bill.

True downtime cost = lost contribution margin + idle labor + recovery overtime + expedited parts and service + rework or material loss + the cost of late orders

Not every outage creates every one of those costs. But if the financial model only includes the technician’s invoice, it is probably missing most of the damage the production team actually feels.

After that, run three operating numbers:

Units at risk per outage = normal finished, accepted output per hour × expected outage hours

Contribution margin at risk = units not produced × contribution margin per accepted unit

Recovery time = units lost during the outage ÷ spare accepted capacity per day

The third number is the one that tends to get ignored.

A facility may technically have enough calendar days left to make up the lost production. But if the machine was already scheduled near full capacity, there is no spare capacity to recover with. The team can work longer or add another shift, assuming the labor, material and downstream packaging are available. Otherwise the missed units stay missed.

That is why a backup should be measured against the contribution margin it protects, not only the number of days it sits idle.

A basic planning formula is:

Backup break-even downtime days = backup equipment cost ÷ contribution margin normally generated per production day

Use your own margin. Do not use retail price, and do not use a number from an equipment vendor’s ROI calculator unless every cost inside it matches your operation.

FlexPak’s current Max Enterprise package illustrates the modular approach. It is built around two standalone Max machines rather than one larger machine and is listed at $34,800. FlexPak publishes up to 49,248 packed-and-closed cones over an eight-hour day with a four- to six-person crew.

Those are up-to packed-and-closed figures under FlexPak’s stated production assumptions. They are not guaranteed accepted output, and they do not automatically include every downstream retail-packaging task.

The point here is not that every facility should buy that package. The point is that capacity is added in pieces. A company can start with one cell, add plates and workflow capacity, then add another machine when the order book supports it. If demand falls six months later, the operator is not forced to keep one oversized line running just to justify the check that was written for it.

That matters in cannabis, where a 500,000-unit month can be followed by a 200,000-unit month with very little concern for anyone’s depreciation schedule.

There are operations where one fully automated line is still the right answer. Stable demand, consistent material, a narrow SKU mix, trained in-house maintenance, critical spare parts, responsive support and enough finished inventory to cover a reasonable outage can make the economics work very well.

But “the machine is reliable” is not a continuity plan.

Before putting most of the production volume onto one unit, ask:

• What is the published daily and weekly maintenance requirement?
• What are the five most common stops, and which can our own team clear?
• Which parts fail or wear most often, and which of them will be stocked on-site?
• How quickly can remote support respond? When is an on-site technician required?
• Where are the parts stocked, and what is the realistic delivery time to our facility?
• What percentage of normal capacity remains if this machine is offline for a full shift?
• Is there enough unused capacity elsewhere to recover the lost units?
• Do our supposedly independent machines still share one grinder, closer, compressor, circuit or trained operator?
• How many finished days of inventory are we willing to carry as a buffer?
• What will one hour, one shift and three days of downtime cost in contribution margin and delayed orders?

Then decide how much redundancy is worth.

You do not build backup capacity because you expect the equipment to fail every Tuesday. You build it because the order still has a ship date when it does.

The best production plan is not the one that never stops.

It is the one that does not stop everywhere at once.

Part 10 is next: Questions to Ask Before Buying. We are going to pull the entire series into one practical equipment-buying checklist.

“

The best production plan is not the one that never stops. It is the one that does not stop everywhere at once.