The Need for Efficiency and the Changing Landscape
The world population is currently made up of about 7.5 billion people, which is about twice as many people as there were as recently as 1960, and this is projected to reach 9.7 billion by 2050. With the rapidly increasing population, more food needs to be produced from the same amount of land as there has always been, meaning there is a constant pressure for the agriculture industry to increase efficiency of food production. Over the last 20 years, there has been a major shift in technology applied to agriculture that has significantly increased efficiency, with greater output (crop yields) produced for less inputs (seed/fertilizer) per acre of land. GPS technology has facilitated precision agriculture techniques and technology that have been a large factor into increasing efficiency, with functions such as auto-steer, variable rate application, prescription application, overlap control, and many others. This has enabled the ability to apply the right amount of inputs based on soil and moisture conditions in the right place to optimize yields. While GPS technology has now become well established, there have been major advances recently in Internet of Things (IoT), machine automation, and data management (big data) technology that are now ready to come together to make autonomous machine operation in agriculture a reality. The path towards highly automated and autonomous machines in agriculture is complex, but will evolve quickly due to the real and demonstrable value that it offers. Farmers are known to be willing innovators to try new technologies that promise new efficiency and increased profitability, but this needs to be proven and reliable to be accepted. As well, new technology must work with existing infrastructure and machines, as the move to autonomy will happen over time.
The State of Technology Leading to Autonomous Machines
Agriculture is positioned better than any other industry to be the first major adopter of autonomous machines for the following reasons:
- Precision agriculture technology based on GPS have laid a foundation of precision control in the industry.
- There is a solid business case for real value and increased profitability in agriculture with autonomous machines, it is not just a novelty.
- The current high cost of equipment and shortage of skilled machine operators makes autonomous systems an economic option.
There are three major areas of technology that have developed significantly in the last few years that are now ready to come together to make autonomous machines in agriculture a reality. An overview of each of these is provided below.
Internet of Things (IoT) and Connectivity Technology
IoT is a term that is used to describe embedded electronics that have Internet connectivity capability. The growth of the Internet over the last 30 years, and the cost-effective electronics that can be built into every device, have resulted in connected objects for a vast variety of purposes. This can be anything from everyday appliances (coffee makers, fridges), to industrial machinery, to vehicles. IoT technology has resulted in a huge amount of real-time information provided from advanced sensor systems that can be used to make improved business and operational decisions.
In agriculture, this can be seen in the connectivity of mobile machines (tractors, implements) and stationary objects (grain bins, dryers) to provide real-time sensor data giving information about farming operations. There is a unique problem in ag in that Internet connectivity is often not available on the farm. This challenge has been met with local connectivity options that store data locally until it can later be connected to the Internet. Tablets and smartphones have been a good enabler of this technology as they are well suited for local Wi-Fi or Bluetooth connections to store farming information as it happens, and then later uploading to cloud-based Internet servers. Also, solutions for long-range Wi-Fi or other RF communication have grown significantly to build towards a fully connected farm.
The advancement of IoT technologies has also resulted in reliable, robust, and secure ways to connect machines to both local networks and the Internet to allow for real-time sharing of data between machines and remote users paving the way for the remote operation and monitoring needed for autonomous machines.
Data Analytics (Big Data)
With all of the data now available because of IoT technologies and advanced sensors in agriculture, the next challenge is to make sense of it all. While having millions of pieces of live sensor data about farming operations is the first step, if you need to analyze it in a spreadsheet to make sense of it, it is not of much value. The next step is to automate the analysis of this raw data and turn it into information that can be used to make better decisions on farm operations.
Data analytics technologies have grown and evolved rapidly in every industry in the last few years in response to the challenge of managing and making sense of all of this data. There are three main types of data analytics that provide value to end users, which are descriptive analytics, predictive analytics, and prescriptive analytics:
- Descriptive analytics is meeting the challenge of presenting the data that is available, but in a form that can be readily understood. This is really answering the questions Who, What, When, and Where. Examples are showing the actual crop yields for each field on a farm, or tracking the location of grain in carts, bins, and trucks.
- Predictive analytics is meeting the challenge of trying to predict what is going to happen next, so actions can be taken to mitigate challenges, and take advantage of opportunities. Examples are using machine operating data to predict breakdowns before they happen so that preventative maintenance can be scheduled to minimize downtime.
- Prescriptive analytics is meeting the challenge of understanding the Why and How. This is a deeper level of analytics to understand the reasons behind the current situation so actions can be adapted to produce the results desired in the future. Examples of this is the work that agronomists do with a more detailed understanding of growing conditions, setting application prescriptions of seed/fertilizer based on soil and moisture conditions.
Farm management systems have grown rapidly in recent years with the express purpose of meeting the challenge of data analytics in agriculture, but it has been a fragmented market. The tractor manufacturers have developed connected farm solutions, but most of them are closed to work only with a specific brand of equipment. Special function options have been provided by GPS companies for precision farming, but only work with GPS-enabled devices. A third group of companies have focused on overall farm management solutions that are independent of specific equipment, and the challenge for them has been access to machine information, and specifically the information from implements as this is the source of much of the valuable farming information. While none of these provide the complete and integrated solution that is needed by the farmer today, the openness of the farm management solutions presents an opportunity for growth towards a solution. Farmers are pushing for ways to connect 3rd party systems together and need a way to combine the data analytics generated by the control systems that drive their machine functionality with the farm management software. This has resulted in standards such as the ISOBUS FMIS and Task Controller defined standards, but the missing piece is the software and equipment databases, like JCA’s Cumulus Farm API, that links it all together. As autonomous vehicles become more prevalent, it will be essential that these various system components work together and only then can the complete autonomous farm be achieved.
There have been many recent advances in highly automated machines driven from various industries. A major contributor to the technology base has been the automotive industry with a buildup of technology towards autonomous cars and trucks. This has resulted in vision systems, object detection, autonomous driving systems technology that is proven and robust and is ready to be applied to the agriculture industry. However, agriculture has unique needs in this area that differentiate it from the automotive industry. Because of the wide range of farming operations, the challenge for autonomous ag vehicles is not only movement but also automation of the implement function (such as seeding, spraying, harvesting, and grain carts). This requires a merging of precision ag functions that are based on GPS systems as well as many other automated and manual machine movements that may be operator controlled today. There has been a rapidly increasing trend in agriculture implement controls in the last few years that has reached a point where the automation needed for autonomous equipment is ready.
A Vision of the Future
It is clear that today’s technology trends are moving agriculture machinery to higher levels of automation and towards fully autonomous systems in the near future. As autonomous machines reach large production quantities the cost of the machines will drop such that small machines will make more economic sense than large ones, since machines are no longer limited by available operators. These smaller autonomous machines can provide much more flexibility.
Farms in the future will consist of a larger number of smaller autonomous machines that operate in coordinated fashion based on higher-level directives from the farmer. Farmers will be making decisions about the operations of the farm based on real-time data provided from all areas of the farm and advanced data analytics that turn that data into digestible information about the state of the farm. This will bring new levels of efficiency that is needed to be able to feed the world of the future. Farmers will be able to focus on applying their knowledge to direct operations in the best way possible.
While the past decades have focused on the advancement of the tractor, the buildup of technology towards autonomous vehicles and the continued focus and need for precision agriculture shift the focus to the advancement of the implement. These current developments bring a radical shift, questioning the future of the tractor. For the last 100 years or so the tractor has been a key part of farming operations and its existence has been taken for granted. It provides not only the horsepower needed to pull implements though the field and supply them with adequate hydraulic, electrical and in some cases pneumatic services, but they also provide a home for the operator. As a result, the implement manufacturers have had the continuous challenge of trying to make their equipment work within the controls, displays, connections, and power provided by the tractor. However autonomous technology changes this dependency. Autonomous vehicles can work around the clock and a network of vehicles can get the job done concurrently, without the need for additional labour. This means that vehicles can be smaller reducing the reliance on high horsepower and there will of course no longer be a need to carry an operator. So why then is there a need for a tractor? Function specific agricultural machines can be developed which would include their own smaller drivetrains at a lower system cost. One example of this radical shift is the recently released autonomous agricultural machine from DOT Technology Corp (seedotrun.com), with a control system developed by JCA Electronics, showing a vision of farming without a tractor. This is only the beginning of how autonomous machines will change some deep-rooted assumptions about farming and lay a path for new, cost effective technology development going forward.
The Path to Autonomy for Equipment Manufacturers
If manufacturers of agriculture equipment are not nervous of where they fit in the new technology landscape, they are not paying close enough attention. Implement manufacturers have been building more and more “smarts” into the machines they manufacture, which provides more precision control and real-time data that is useful for increasing farm efficiencies. Equipment is evolving with more automation to coordinate more complex operations, more sensors to monitor all aspects of the machine operation, and more connectivity for coordination, control, and analysis. The path towards autonomous systems can happen naturally by adding intelligence to the machine in value-added steps along the way. Implement manufacturers need to be diligent in finding ways that these technologies can be added to the machines they manufacture to add real and immediate value to the farmer, and they need technology partners, like JCA Electronics, that can help them navigate a way forward.