Sorting and portioning of inhomogeneous hemp flowers

Hemp is one of the emerging goods of our time – strictly regulated but completely legal. The controlled requirements in processing require precise working methods. Not an easy task for the project team of the Degree Programmes in Automation Technology at CAMPUS 02.

Cannhelp GmbH is an Austrian company that has already gained a foothold in the emerging health industry of CBD products. The high quality standards require meticulous control – a hitherto manual effort that is passed on to the customer in the form of costs. The sorting of the flowers according to their size plays an important role. The customer approached the project team with the request to develop a concept for an automated sorting of inhomogeneous hemp flowers. The subsequent portioning of the usable flowers should also take place automatically.

In the following months intensive work was invested on every process step. Research and analysis as well as experiments and validation were carried out. In order to find the best possible solution, the composition of the product was also carefully examined: What is the density of a hemp flower and how can it be transported practically without damaging it in the process? For the first transport step, a conveyor belt with a white surface emerged as the most promising procedure after a few experiments. Several cameras and image processing programs were tested for object recognition. Simple systems already provided usable results. However, in order to precisely characterize the flowers, increased programming effort is required. The separation of the flowers with compressed air valves proved to be a very promising approach. Finally, a concept for cost-effective portioning was developed based on a multi-scale system.

This project perfectly illustrates the tasks in automation technology when there is no standardised solution available: The step-by-step logic that has to be experimentally implemented into practice. Amonst others, one challenge was to find a solution for several work steps that is not prone to errors and can also be used with small product quantities.

By using an innovation check for this project, various options for automating the separation and portioning process could be developed. This enabled the project customer to become familiar with digitalisation and automation and significantly increased acceptance for these technologies. The knowledge gained will serve as technical basis for an initial setup of a partially automated system. This is intended to lay the foundation for further process automation and optimisation in the cannhelp company, which will be an important corner stone for further expansion.

For more information, please contact Matthias Primas, Research Assistant at the Degree Programmes Automation Technology at CAMPUS 02, University of Applied Sciences