Some may argue that an ounce of weed is better than a pound of cure, but that is another article for another journal. Today we are here to talk about the secret weapon that is process engineering, quietly and quickly keeping the world from falling apart with “ounces of prevention” that come to life as an interconnected web of high-tech sensors, algorithms, and control panels overseen by hardworking engineers. Process Engineering serves the first line of defense in making sure that everything you obtain is as safe and reliable as possible.
In nearly every regulated industry, there is required testing for product usually regulated by a governmental agency and its affiliates, some examples include: Fuel Octane levels with the EPA; Pharmaceutical formulation through the FDA; and food and beverage integrity through the USDA. However, this third-party testing is not the majority of testing done during manufacturing. In fact, it may comprise the smallest percentage by type, since most products undergo dozens of rounds of internal testing prior to being submitted for external scrutiny.
Why?… Well, in the immortal words of Don Henley: “People love it when you lose, they love dirty laundry”
Photo courtesy of Sage Analytics
Failed tests could mean revoked licenses, a threat to product quality, or potentially, a leaked story of contamination, even if there is no public risk and the contaminated or inappropriately formulated product never gets released. If that bad batch of product gets loose, it means a recall at best, and a risk to consumer and public safety at the worst.
That said, if a bad batch of product never gets released and is quietly and appropriately dealt with in house, it never becomes a product. Ingredients can often be reformulated, cleaned, scrubbed or modified at critical control points along the production cycle before significant losses occur. Here is where the real value comes back into play as we reference the great Benjamin Franklin’s “ounce of prevention”. By implementing in-house measurement, we can prevent problems before we ever create them, let alone have to fix them. In engineering, this thought pattern has given birth to QbD (Quality by Design), and CGMP (Current Good Manufacturing Process), among other types of design and regulation paradigms. Therefore, measuring at every step along the way has become a significant aspect of modern manufacturing.
Following the manufacturing process of cannabis from purchased trim to manufactured branded products, there are several easy targets that could eliminate variance in output from the beginning, by measuring the following:
- Raw ingredients before buying, checking specification within accepted and agreed upon range
- Raw ingredients for degradation after sitting on shelf at time of use.
- Mixing cannabis to obtain homogeneity of potency for process input.
- Mixing chemicals for proper chemical reactions (solvent saturation, baking, decaffeinating, dosing).
- Checking and weight mixing intermediary products between processes for optimal yield.
- Adding excipients (inactive ingredients), preservatives, and supplements with quantitative dosing.
- Checking final product immediately prior to packaging.
- Keeping retention samples of each batch to measure degradation via shelf-life studies.
- Cleaning validation of equipment between batches.
Considering all these needs that must be met for regular product cycles, how are other companies conducting this testing while still profiting on regular products like bread (consider $2/ pound bread rather than the $1300+ pound of cannabis)? The answer is process analytical technologies (PAT), often connected, in-line or at-line sensors that give real time feedback to automation equipment that tells factories to take action or stop until the problem is fixed.
Everything from the caloric content in your store-bought bread, to the octane in your gasoline, and the dosage in your prescription medications is carefully metered by instruments at every critical step along the production cycle. These may be obvious at first, but there are other applications such as the opacity of plastic used in containers, or the moisture and color in dog treats or kitty litter. Many products are also required to be tested for adulterants and heavy metals.
Assuming you are intrigued, you may be wondering how this compares and contrasts to traditional wet chemistry laboratory analyses, or even what makes a piece of tech optimal for given type of test. People often ask me, incredulously: “If in-house testing works, why even use complex equipment in third party labs? Why can’t I just test my cannabis in-house and be done with it?” To resolve this apparent conflict, we need to look into the different nature tests and use cases of the equipment.
In the wet chemistry lab, many of the different types of required tests can be carried out by a limited number of expensive and complex instrument types (coupled chromatography and mass spectrometry for content and adulterants, PCR for biologicals, and ICP MS for heavy metals). To achieve the best chance at identifying even the smallest levels of adulterants, these machines are designed to allow for the discrimination and separation of chemical species by mass or charge, coupled with the highest detector sensitivity for quantification of these tiny quantities. However, the cost for that sensitivity is that the sample must be destroyed and the output of these tools has a rate of minutes per sample, often allowing less than 200 samples per day even assuming 24-hour operation. For greater testing frequency needed to measure large volumes of variable material, the desired throughput (data generation rate) doesn't allow for minutes per test. Wet lab tests would also ruin large amounts of material, and the cost of testing would become an onerous burden, in addition creating an excessive turn-around time. Sound familiar? On the other hand, if your future depends on the reliability of a single test, you would want it run on the fanciest piece of equipment possible before results go to the general public or a licensing agency.
Being able to make quick decisions about day to day production, and making sure the final product is perfect are both critical to success in cultivation and manufacturing, arguably even more so in the heavily scrutinized cannabis industry. Instead of choosing between perfection and data volume, other regulated industries meet both goals through a compromise. It is generally structured that the in-line measurements are designed to be non-destructive and blindingly quick to obtain data volume and statistics for homogeneity, trends, and process control with hundreds to thousands of data points every hour. Wet lab tools are reserved for the stringent final product testing requirements, as well as a handful of tests that confirm the performance of in-house tools (often at a ratio 1000:1 or 10000:1).
The tradeoff is that in-line tools inherently lack the species separation components of wet-lab tests (e.g. chromatographic columns), and therefore, use modeling and multivariate analysis machine learning techniques to compensate and ensure agreeance to wet lab tools. Even then, since these inline tools are inherently modeled off of the wet lab tools, by definition, they cannot have the same accuracy and precision as the best tools with the best chemists under optimal environments. Nevertheless, thousands of data points with excellent accuracy can help more so than a dozen data points that are “perfect”. The trick is to apply the right type of test to the right testing paradigm.
In-house testing is excellent whenever a large volume of data can be used to monitor your business: harvest optimization, selecting vendor product, negotiating pricing, testing and weight mixing trim to ensure proper regulated input for an extraction, generation and monitoring of standard operating procedures for a product formulation, or even potency-based dosing into edibles to ensure batch consistency.
The cannabis industry deserves the same. Safe cannabis is the only way towards any chance of national or even global legalization, and the best way to safe cannabis is measuring early and often. Measuring in-house means finding and eliminating sources of error and basing both economic and scientific decisions on large volumes of data without having to move slower. Whether you are cultivating or extracting, or formulating edible or inhalable products, control is critical to brand reliability, and in-house testing offers control without breaking your budget. We make these tools at Sage Analytics, backed by 19 years of experience successfully making the same tools for the pharmaceutical industry. When you are ready to have every detail and every aspect of your process under strict control, call us…