Leveraging Similarities in Smart Manufacturing and Smart Agriculture

I recently participated in a workshop on Controlled Environment Agriculture, and I believe we can learn from an exchange of lessons learned between Smart Farm and Smart Factory initiatives. Both initiatives have been making good technology implementation advances with industry leaders and large companies. However, on both initiatives, we need to do better with small-medium size companies which make up 90% of the businesses in both agriculture and manufacturing.

Both Smart Farm and Smart Factory initiatives share the potential to revolutionize the industry with benefits that go beyond reduction of cost and improvement of quality. Both share benefits in labor efficiency through automation techniques, and energy efficiency through process optimization techniques. Both benefit the whole ecosystem by increasing speed and coordination among partners in the supply chain, and both benefit their communities by providing high-tech, well-paid jobs.

Greenhouse automated Smart Farms can have benefits for the environment including reduced use of land, water, and fertilizers, and benefits for the food system like enhanced food quality and fresh produce in places where it is not available today.

The Smart Farm use cases include monitoring of crops and livestock, and farming with drones and autonomous machines. These use cases feel very different than the use cases we see at the factory. However, when we look closer at the Smart Factory use cases, we see several that are very applicable to the Smart Farm. For example, we can use similar visualization, automation, predictive maintenance, mobile platforms, and optimization techniques on both sides.?

There has been big innovation in sensor and communication technology allowing Smart Farms to collect all types of real-time data directly from the field about the growth and well-being of the plants. Real-time data is critical to smart techniques, but raw data has limited applications until it is coupled with additional contextual data like the specific crop, soil amendments, equipment, and season, and organized by information models for enhanced analytics and insights; insights that are used to manage the farm and intervene as soon as possible to control and enhance the growing process.

The combination of technologies for sensing, information models, artificial intelligence (AI), workflow, and controls to create a smart platform that enables smart techniques for future smart farms that are more resilient to climate, location, and market disruptions. A smart platform with these capabilities is needed for both the Smart Farm and Smart Factory. The sensor and apps of the Smart Factory might look very different, but the platform that turns data into insights and connects technologies and systems together can share a lot in common.

Common Technologies and Capabilities

The technology core that is common to both Smart Farm and Smart Factory includes: ?

• Advanced automation and robotics
• Smart sensors and IIoT connectivity ?
• Mobile, edge and cloud computing
• Advanced analytics and artificial intelligence

But the common core is not just about technology. It is about how we thread these technologies together to create smart capabilities for the business; capabilities that include:

• Collecting data
• Connecting things
• Providing insights
• Automating flow and control
• Augmenting the workforce
• Coordinating the value chain

Cloud and Edge Computing

Smart Farms need to connect crop monitoring devices, autonomous farming equipment, workers in the field, and management offices. Cloud computing allows access to systems and data from anywhere in the field to the office as long as we have an internet connection.

Cloud computing brings scalability into the systems architecture by providing Infrastructure-as-a-Service (IaaS) and Software-as-a-Service (SaaS) that scales when you need more computing power. It can also provide redundancy and backup services. The pay-as-you-go pricing model allows businesses to increase their computing power as they grow.

Automation of industrial systems can be done in the cloud or at the edge depending on the tolerance for latency in the specific process you are controlling. Edge computing is co-located on-premise closer to the source of data and can be used to aggregate, filter and send less data to the cloud. Edge computing helps save bandwidth, reduce latency, and enhance scalability by distributing the analytical load to process IIoT data.

Information Platform

An important part of turning data into insights is the use of information models. Information models add structured context for enhanced decisions, correlation, and identification of opportunities for improvement in the overall production process.

CESMII has a technology platform available to develop, demonstrate and test information models on open standards like OPC UA. The platform is available for production use and is also used as the backbone for R&D projects with CESMII members. The data organized in the information models becomes available through a GraphQL open API to modular applications for visualization, analytics, workflow, and operations management.

CESMII is collaborating with leading organizations on industrial data standards in the US and Germany including Platform Industrie 4.0, VDMA, and OPC Foundation to promote the use of standards, grow the libraries of open standards, and make them easier to use for plug-and-play connections between assets, processes, platforms, apps, and systems. Hundreds of information models are already available, and the number continues to grow. Any new SM project should start by researching the library before deciding to create a custom information model from scratch.

In addition to an information model library, CESMII is growing a library of smart manufacturing Apps and Bots that mine the information models and provide modular functions for manufacturers including advanced control and optimization capabilities that use artificial intelligence techniques. The CESMII ecosystem is currently focused on smart manufacturing use cases for the libraries, but smart agriculture initiatives could benefit from similar capabilities. It would be great to get experts involved from that ecosystem to expand the open libraries.

Advanced Process Control and Optimization

Production processes can be very complex with many control parameters and environmental conditions affecting production in many ways. Optimization approaches provide a methodology to adjust the production process control variables based on different running conditions and decision models to deliver a higher level of performance, quality, and energy consumption.

Advanced process control and optimization is what truly makes a system “smart” by going beyond the reporting dashboards and establishing proactive practices that use artificial intelligence (AI) to automate routine decisions and trigger action on alarming trends or non-routine situations.

The CESMII ecosystem has worked on many R&D projects that developed AI algorithms to optimize processes and save energy in multiple industries. For example, in one project with the Pacific Northwest hop industry, Oregon State University and Ectron used sensors to monitor temperature and humidity and applied developed AI models to reduce hop drying times by 14% and reduce energy usage by 10%. Another project with General Mills and ThinkIQ developed AI predictive analytics to optimize energy usage and achieved 5% reduction in energy waste at several plants.

Another project with a food manufacturer used CESMII’s SM platform to ensure that their product could be labeled gluten free with certainty. Oats, wheat, rye, and barley look very similar in bulk. The specific grain and containers must be carefully tracked to avoid contamination of containers in the flow of oats through the silos, trains and trucks used to distribute grain from the multiple farms to the factories. This used to be a highly manual process tracked through blackboards and spreadsheets. Now it is automated, and the process minimizes the waste caused by contaminated batches.

New Skills in for the Workforce

One more common element between smart manufacturing and agriculture is the need for new skills in the workforce to implement and utilize these new technologies and techniques. New skills include:

• Integrating IT and OT systems leveraging edge and cloud computing??
• Data platforms that handle big data with technology like data lakes
• Newer connectivity techniques like MQTT, OPC UA, GraphQL
• Programming with Python instead of ladder logic
• Developing Apps with low-code platforms

CESMII has many education partners providing Smart Manufacturing education for the workforce. There is a wide variety of programs for different skills within SM and formats from on-demand courses that are 45min, to instructor-led certificate programs that are several weeks and are available remotely or in-person. You can learn more about these programs through the education catalog at www.cesmii.org.?