Redefining Hemp Stalk Processing Economics with AI

Robert Czinner, Canadian Industrial Hemp Corporation -
hemp plant automation
Illustration: © IoT For All

According to the Oxford Dictionary, “optimization is the action of making the best or most effective use of a situation or resource.” However, this is not always what happens in an optimized manufacturing environment. That is because, during the optimization process, the “best” solution may need to consider one or more additional constraints, which may alone determine the final determination, such as completing a scheduled Open Order.

For example, an optimization focus on “Production Cost” must have the ability to automatically and simultaneously factor in & correlate other interactive operational constraints such as the production process, productivity, product output values, shipping logistics/costs, production scheduling, cash flow requirements, and open orders.

To process these considerations simultaneously requires real-time data collection and communications that only digital technologies can deliver: It is simply beyond the scope of human beings. Essentially, real-time optimization requires the dynamic monitoring, control, and synchronization capabilities that only AI can provide.

Hemp Stalk Processing

The proven large-scale commercial technology and Business Models in today’s European hemp stalk processing (“decortication”) industry today were initially developed in the 19th century, with some automation upgrades being introduced over the last 20 years. Hemp stalk processing only converts hemp stalk into its 2 component fibers – bast and hurd – both low-value commodity outputs. To achieve large, consistent volumes of quality-controlled outputs requires a significant capital investment, manageable agronomics, and relatively high labor costs. It is hard to generate free cash flow without producing large volumes and engaging in vertical integration.

Opportunity

What the decortication industry needs is a way to redefine its economics of processing: It needs to improve its:

  • Ability to output large volumes cost effectively
  • Production flexibility
  • Overall productivity
  • Value Added
  • EBITDA

Fundamentally, what is needed are tools to facilitate real-time digital communication and integration between a wide range of digital devices. Additionally, having an integrated system architecture enables the physical integration of primary processing and secondary production – improving productivity and further reducing costs.

The opportunity is to start at the beginning of the production process, digitally track all the constraints, and be able to review their direct and indirect impact on throughput, quality, and consistency, in real-time through the processing and production process.  The data can then be analyzed by AI’s inherent Machine learning capabilities to determine the constraints of its optimized solution.

Artificial Intelligence

AI introduces the unique ability to monitor, control, and synchronize over 10,000 digital systems operating in real-time – simultaneously.

AI offers a unique approach to searching for, identifying, and evaluating data and then analyzing the results to determine a specific solution. And it does so at lightning-fast speeds. This unique ability allows AI to assess and analyze digital data in such a manner as to predict the “most likely” optimization outcome – and the course needed to be followed to get there.

Artificial Intelligence (AI) plays many critical roles in manufacturing today. It is intrinsically connected with Industrial IoT (IIoT) and drives Industry 4.0.

Essentially, any digital optimization solution being pursued sets out to identify and correlate all the data relevant to illuminating the sought result. This 2-part process results in the appropriate data, which will be “optimized” to achieve a specific “best” solution.

Machine Learning

Processing and production facilities generally rely on AI to utilize its unique ML capabilities capable of integrating vast volumes of real-time data, analyzing it, and then providing the profound insights and predictions needed to help drive better decision-making throughout the organization.

Process manufacturing is a highly competitive sector, with swift-changing markets and complex systems that have many moving parts: To drive innovation and improve profitability, process plant operators need a Sustainable Competitive Advantage to differentiate them in the marketplace.  ML in manufacturing powers the unparalleled benefits of predictive analytics, robotics, predictive maintenance, and automated processes: All of which help make operations more efficient, profitable, and safe.

Some of the direct benefits of Machine Learning in manufacturing include reducing common, repetitive process-driven losses generated by:

  • too much labor cost
  • Too little throughput – inadequate scale
  • Wasted materials
  • Inconsistent and inadequate quality

In addition, the capacity is inevitably by optimizing the production process. Inevitably, growth and the expansion of product lines are facilitated by the improved cost and potential revenue dynamics.

According to a 2020 PwC report (Digital Factories 2020: Shaping the future of manufacturing), the adoption of ML and analytics to improve predictive maintenance is predicted to increase 38% by 2025. Analytics and ML-driven process and quality optimization are predicted to grow 35%, and process visualization and automation, 34%. PwC sees the integration of analytics, APIs, and big data contributing to a 31% growth rate for connected factories in the next five years. The summary below demonstrates how significant Senior Management sees digital connectivity playing in the Manufacturing domain.

Machine Learning is the production environment for optimization.

Optimization

Optimization can currently be applied both to product design as well as process design. In fact, not only do you apply optimization to design the best manufacturing process, but you also use the optimization process to make the best use of your holiday time!

You apply optimization to design the best cars, equipment, airplanes, cell phones, and everything in between. Optimization in hemp stalk and fiber processing defines a meaningful financial opportunity: Consistent, high-quality bast fibers for textile applications can be worth up to $2500 per ton, whereas lower quality fiber used for animal bedding is usually sold for no more than $600/ tonne!

Optimizing Hemp Stalk Processing and Fibers

The Production Process: Flexibility & Customization

The ability to automatically quality control the hemp stalk before it is processed makes a huge difference in the value of the outcome. The logic is basic: Process only bales containing high-quality stalk into high-value end products, and bales containing low-quality stalk into lower value products.

Conventional stalk processing inevitably results in guessing what quality the stalk is INSIDE the bale. In this regard, specific fibers can be identified by the System and sent directly to specific secondary production lines.

Productivity

The ability to monitor all pumps, valves, motors and sensors enable the System to utilize an Automated Predictive Maintenance System – while also tracking the data underlying the operation of every device in each facility. Machine Learning will teach the System how to operate more time efficiently – automatically improving productivity over time.  This is a critical consideration for hemp stalk processing which has low productivity.

Product Output Values

The different potential output values for hemp fiber play a key role in optimizing the value of output!  Knowing what can be made from incremental portions of large volumes of a natural product – such as hemp – in advance of producing it will result in higher quality, less waste, and much greater overall productivity.

Shipping Logistics and Costs

In today’s North American reality, shipping costs are material: They impact overall cost structures. The ability to automatically correlate expected delivery dates and locations could impact overall operating margins.

Production Scheduling

Production scheduling is an art as well as a science. It involves coordinating input supplies with output orders and being able to project how long each production order should take to process, produce, and package.  The System automatically determines deep insight into all the constraints and considerations that may impact the predicted results using historical data.

Cash Flow Requirements

It is not unusual in a production business to consider cashflow realities when planning production: Month-end financial commitments sometimes result in faster-paying customers getting preferential production. An AI-driven System can automatically monitor each Customer Account and factor in its $ size, payment in average days, and the Profit per Tonne they provide: The System can be taught to factor these constraints into its decision-making as well.

Open Orders

Ensuring that production schedules and deliveries are fulfilled as promised is a form of optimization – focused on Customer Satisfaction.

Better ROI

Optimization generates better ROI and better rewards. The ability to optimize hemp stalk decortication redefines the economics of producing hemp fiber. It thereby redefines the size of the markets and product opportunities for which hemp fiber can be used. Optimization will be critical to help bring hemp prices down to compete more effectively with existing feedstocks, for which hemp could be a preferred substitute.

Author
Robert Czinner, Canadian Industrial Hemp Corporation - Founder and CEO, Canadian Industrial Hemp Corporation

Contributors
Guest Writer
Guest Writer
Guest writers are IoT experts and enthusiasts interested in sharing their insights with the IoT industry through IoT For All.
Guest writers are IoT experts and enthusiasts interested in sharing their insights with the IoT industry through IoT For All.