Awakening to the potential of artificial intelligence in coffee
Image courtesy of Avocado Studio & Makr Shakr
Although many initiatives are still in test stages, artificial intelligence can help to advance the coffee industry. With artificial intelligence, solutions can be incorporated throughout the entire supply chain, including coffee grading, identifying diseases and even in the roasting process. By Anne-Marie Hardie
Today’s companies are driven by data. This data provides industries the information needed to resolve challenges, increase productivity, streamline operations and increase profit.
Devastating environmental situations like coffee leaf rust prompted the industry to work together to collect and store data to see if there were able to identify risk factors and best prevention strategies. However, collecting, inputting, and analysing this data requires significant human resources. Artificial intelligence-driven solutions can help alleviate this strain on resources reducing human error while also allowing experts to focus on unique situations that could take their business to the next level.
In fact, computer science researchers Alexandre Pereira Marcos, Natan Luis Silva Rodovalho, and André R Backes at The Federal University of Uberlandia, Brazil, are currently training a Convolutional Neural Network (CNN) to identify coffee leaf rust. The spray robot would use image processing and machine learning to identify which leaves are contaminated with coffee leaf rust, and in turn, isolate which plants need to be treated. The results are promising. The current model was able to detect the leaves that were impacted by coffee leaf rust, but also reported that the areas surrounding the leaves were infected.
Although it was a broader response than they had wanted, the researchers emphasised that the detection was accurate as identifying the central region of infection was their main goal. Future research in this area will help to refine the CNN models with the goal of developing a robot that can be used at the farm level.
Transforming the Farmer Experience
“Artificial intelligence allows us to focus our time making decisions that we are really good at instead of focusing it on areas that are consistent, and in turn, low value,” said DJ Bodden, chief operations officer, Farmer Connect, Vernier, Switzerland. “It allows humans to do what they do best – adapt to unique situations.” Through the integration of artificial intelligence, data can be collaborated and collected to identify key trends, markers to pay attention to, and in turn, free individuals to focus on the more challenging tasks.
One of the key challenges that Farmer Connect wanted to address was to provide a tool that would connect farmers to their supply chain to improve transparency and their overall livelihood. Powered by IBM, the Farmer ID application provides a tool that will aid farmers with their day to day business, while also providing an identity wallet. Their goal was to develop a blockchain solution that enabled the user to feel secure and in control of the data that is shared. The system allows the user to select which data they choose to share in this platform, and which they keep private. Providing users with this option makes the software much less intimidating to adopt. At the same time, the data that is entered into the system can be used to help identify trends, challenges, and increase overall efficiencies.
The blockchain technology enables farmers to connect with their entire supply chain, providing a digital record of their purchase transaction. “Now, when the farmer goes to the bank, there is a record that validates the purchase price of the coffee making it easier to obtain loans,” said Bodden. The system also provides the ability to use QR codes in the retail space providing consumers with access to the story of the coffee farmer including a product page, photographs and videos.
Connecting Across the Supply Chain
When it comes to coffee, Ankita Valeja, manager, market research at CropIn, Karnataka, India, emphasised that smart technology is lucrative for the entire industry. These systems provide companies with decisions that are based on real-time data maximising outputs and increasing efficiency in production. “The CropIn solution provides traceability, which brings transparency,” said Valeja. “These technologies provide early warning systems while also using weather data in conjunction with historical disease occurrence to better predict the likelihood of a disease outbreak for specific crops.”
Founded by Krishna Kumar, CropIn was designed to aid the farming community by identifying which areas may be more susceptible to a specific disease and help anticipate its potential impact on the next harvest. By monitoring their risks, farmers can make proactive, planned decisions.
Once processed, CropIn’s mWarehouse system steps in to help strengthen the link between the growers and the remainder of the supply chain. “The application facilitates efficient management of packhouse processes with its flexible inventory management and SKU tagging, thus cutting down the number of hours spent in manual labour and significantly increases productivity,” said Valeja. Suppliers are equipped with a digital record of the production and distribution history, allowing them to quickly identify any challenges in the supply chain, reducing losses and improving efficiencies.
The three-tiered system, SmartFarm, SmartRisk and mWarehouse, provides users with an artificial intelligence solution that moves across the supply chain. “Coffee farmers in all developing nations are primarily smallholder farmers who engage with large agribusinesses and buyers,” said Valeja. “Implementing a comprehensive system increases transparency and visibility over the entire process from farm to shelf.”
Green coffee importer Sustainable Harvest will be leveraging CropIn’s platform to ascertain the origin of coffee marked as cultivated by women, by recording the details of individual growers and allowing end-users to trace this information back from the package’s label, and realise the impact that women coffee growers have on the coffee industry.
“Artificial intelligence can find solutions across the entire supply chain from identifying the best practices for collection to looking at the transportation system, including providing strategies that minimize carbon emissions and improve resiliency,” said Tatiana Kalganova, post-graduate research director, Brunel University, United Kingdom.
Brunel University worked in collaboration with Caterpillar Logistics Research and Innovation team to develop an AI algorithm that helps resolve some of the common challenges in transportation including shipping half-full containers, route optimisation, and reduced the environmental footprint. “Typically supply chains are built where everything goes through one major hub, and if something goes wrong with that hub all business gets stalled,” said Kalganova. “Artificial intelligence can help ensure that we are not reliant on one hub.”
The AI algorithm considered several factors including energy price volatility, ocean lane cost commitment, the carrying cost of in-transit inventory and tariff effects. This solution has been successfully deployed in Caterpillar’s transportation network.
At the farmer level, Brunel University recently developed a “magic bean” solution that uses sensors to collect data about the soil and growing conditions to provide solutions to farmers that will increase productivity while decreasing costs. “The system uses artificial intelligence to provide localised feedback on the conditions of the soil, humidity levels and temperature, allowing farmers to better allocate their resources,” said Kalganova.
Taking Artificial Intelligence into the Retail Experience
Artificial intelligence is not just for the farmer. This technology has transcended into the retail experience, including enhancing the customer experience through tailored communication and promotions to delivering a barista level drink to the consumer.
“Drinking coffee has become an integral part of the daily rituals for millions worldwide,” stated Emanuele Rossetti, CEO, Makr Shakr, Turin, Italy. “At Makr Shakr, we think robotics can make this experience faster, avoiding lines, counter confusion and misspelled names!”
The robotic bartenders have top-quality coffee machines integrated in their systems that are able to produce more than 100 cups of coffee (or any milk-based beverages) in an hour. It is also possible for customers to place their orders in advance, including saving their favourite orders for a faster transaction next time.
One of the core challenges with the café experience is determining the next beverage trends. To help respond to this dilemma, Signals Analytics developed a platform that will aid the food and beverage industry with predicting trends and insights into the industry. “Signals Analytics alleviates the challenges food and beverage brands face in making strategic product decisions by providing smart, holistic, timely, and actionable market intelligence gleaned from a multitude of external data sources,” said Gil Sadeh, CEO of Signals Analytics, Vienna, Virginia.
The platform integrates industry expertise with smart contextual engines, to provide users with a holistic view of the entire food and beverage landscape, including the relationship between different categories and trends.
These solutions are only a small fraction of how artificial intelligence can help to evolve and advance the coffee industry. Artificial intelligence solutions can be integrated throughout the entire supply chain, including coffee grading, identifying diseases, and even in the roasting process. However, to get there, the industry needs to work together to collect the data and define the parameters that will enable the more common challenges to be coded.
“We need to know that the jobs that we are doing today may not be the jobs that we are doing tomorrow,” Bodden explained. “Artificial intelligence does not displace jobs – it creates new ones, allowing people to focus on the aspects that make a difference.”
- Anne-Marie Hardie is a freelance writer, professor and speaker based in Barrie, Ontario. She may be reached at: [email protected].