Tuesday, January 24, 2017

Cannabis automation

Like George, I also intend to join the cannabis industry post-graduation, and like him, seeing Thibaut's presentation took my mind straight to his technology's potential in the cannabis industry, though in a slightly different direction.

In terms of cannabis production, I see the key challenge being the standardization of the product.  Why? To date, the interactions between the human body and the active compounds in the cannabis plant remain relatively poorly understood.  Different people's bodies respond differently to the same dosage of cannabis, taken from the same flower.  To complicate things further, the same person will have different responses to the same dosage of cannabis of the same strain -- even from the same production batch -- if taken from a different plant.  This is due to the "entourage effect," whereby the body's response to cannabis is dependent on the relative ratios of over 100 cannabinoid compounds.  That's because the plant's production of the various cannabinoid compounds is extremely sensitive to environmental factors, in ways that are still somewhat mysterious.  I believe that the consumer (particularly at the entry level) demands a repeatable experience, which requires the producer to market a highly standardized, six-sigma product.  For the producer, this is a major challenge.

Now, on the one hand, the array of sensors and data that Fruition has, could dramatically enhance the industry's scientific understanding of the relationship between time, environmental factors and the chemical profile of the resulting plant material.  To the extent that cultivators could apply this research, it is easy to imagine plants being grown much more effectively to generate a specific desired therapeutic or recreational effect.

On the other hand, the same technology could be applied toward automating a production facility to ensure that each plant grows in a virtually identical environment.  The same technology Fruition uses to determine the nutrient needs of vines can also be used to detect issues in cannabis plants, and flag them for the grower.  Indeed, much of the expertise needed to grow revolves around learning to identify a range of conditions.  Automating detection would enable the cultivator to rely on a lower skilled work force (driving down costs) while enhancing the quality of the end product (potentially enhancing consumer willingness to pay.

4 comments:

  1. Conrad - I thought Thibault's comparison of the wine industry to the cannabis industry to be super interesting as well. I wonder to what extent cannabis might stratify into the types of product and purchaser that wine does. For instance, could a cannabis company look something like Robert Mondavi in having the bulk of its sales from a low-priced, accessible brand like Woodbridge; a more mid-priced house brand; and a premium wine like Opus One?

    ReplyDelete
  2. Really interesting post, Conrad. After reading this you prompted me to write my own post on similarities/differences to the almond industry.

    ReplyDelete
  3. This is a great post - I certainly didn't appreciate how a lack of scientific understanding of cannibis hampered its development, and how this technology could not only improve agricultural productivity, but also enhance the value of cannibis for consumers.

    ReplyDelete
  4. I'm surprised to hear about the challenges in quantifying potency of marijuana, particularly since its been medically prescribed for some years. A cursory google search* indicates that medical dosage is measured by % of THC.
    Your post makes me think that this is probably insufficient, given the presence of "entourage effects." Do you have any ideas of what a better metric for dose/potency may be going forward?

    ReplyDelete