Agriculture 4.0 is emerging and requires an expended sensor suite. Agriculture 4.0 refers to systems that employ drones, robotics, the Internet of Things (IoT), vertical farms, artificial intelligence (AI), renewable energy, and advanced sensor methodologies. Agriculture 4.0 is similar to Industry 4.0 in some ways: Where Industry 4.0 is designed to support automation and mass customization of production processes, Agriculture 4.0 is expected to support autonomous operations and mass customization of farming practices across microenvironments.
Integrating digital technologies into farming, agricultural operations can target resources needed to increase yields, reduce costs, and minimize crop damage, water, fuel, and fertilizer usage. This FAQ looks at sensor technologies under development for Agriculture 4.0, including wearables for plants and hyperspectral imaging, the EU’s Agricultural Interoperability and Analysis System program, and the security challenges related to wireless sensor networks and the Internet of Things in Agriculture 4.0.
Wearables for plants
Graphene and fiber optics are two technologies used to develop wearable sensors for plants. Graphene sensors can measure the time it takes for different crops to move water from the roots to the lower and upper leaves. Initially, researchers are using these sensors to help develop plants that use water more efficiently. In the longer term, these graphene sensors on tape (also referred to as ‘plant tattoos’) are expected to support the design of inexpensive, high-performance sensors for Agriculture 4.0 applications (Figure 1) and help improve the efficiency of irrigation systems. The process used to make the sensors can produce devices that are several millimeters across with features as small as 5 μm. The small feature sizes increase the sensitivity of the sensors. The conductivity of the graphene oxide in these sensors changes in the presence of water vapor, enabling the measurement of transpiration (the release of water vapor) from a leaf.
Figure 1: Graphene on tape can be used to fabricate wearable sensors for plants. (Image: Iowa State University)
Fiber Bragg grating (FBG) sensing technology is also being developed for agricultural applications. An FBG acts as a notch filter that reflects a narrow portion of light centered around the Bragg wavelength (λB) when illuminated by a broad light spectrum. It’s fabricated as a microstructure inscribed into the core of an optical fiber. Unlike graphene sensors, which are an emerging technology, FBG sensors are already in use in several areas, including aerospace, civil engineering, and human health monitoring. FBG sensors can be fabricated with high sensitivities, small size, and lightweight. In the case of agricultural sensors, the intrinsic sensitivity to strain (ε), and temperature variations (ΔT) of FBG technology are being combined with a moisture-activated polymer to detect relative humidity changes (ΔRH) in the surrounding air. In addition, FBG sensors can be multiplexed to support monitoring of both plant growth and environmental conditions in a single device. The FBG designed for agricultural applications consists of three segments, one for ε sensing, one for ΔRH monitoring, and a third optimized for ΔT measurements. It was fabricated using a commercial FBG with a grating length of 10 mm, λB of 1533 nm with a stretchable acrylate coating. The coating protects the FBG and improves its adherence to the plant’s stem.
From multispectral to hyperspectral imaging
Multispectral imaging is an established agricultural sensing technology. It can detect subtle changes in plant health before visible symptoms are apparent. For example, a drop in a plant’s chlorophyll content can be detected before the leaves are visibly yellow. Multispectral sensors use the 712 to 722 nm wavelengths (the red edge band) where indications of stress are most easily identified. Multispectral imaging can be implemented using fixed installations where the sensors travel back and forth on a track system in a greenhouse or across an open field. They are also well suited to be carried aloft on a drone. For example, in one configuration, a drone-based multispectral imaging system can scan a 100-acre field (at 400 feet above the found with a 70% overlap) in less than 30 minutes (Figure 2). Some of the benefits of multispectral imaging include:
- Early disease detection
- Improved irrigation and water management
- Quicker and more accurate plant counting to optimize fertilizer application and pest control
- Cost reductions from the automation of activities previously performed by walking the fields
Figure 2: Drone-based multispectral cameras can take less than half an hour to scan a 100-acre field. (Image: Coptrz)
The primary difference between today’s multispectral sensors and emerging hyperspectral sensors is the bandwidth (the number of bands and how narrow the bands are) used to represent the data of the electromagnetic spectrum. Multispectral imagery generally uses 3 to 10 bands to cover the relevant spectrum. Hyperspectral imagery consists of hundreds or thousands of narrower bands (10 to 20 nm), providing greater resolution, and covering a broader spectrum range. Spectral resolution, the ability to capture a large number of narrow spectral bands, is an important feature of hyperspectral imaging compared with multispectral imaging. Other advantages of hyperspectral imaging include:
- Higher spatial resolution and the ability to discriminate smaller features,
- Higher temporal resolution and the ability to more quickly sense important environmental changes such as the need for irrigation
- Higher radiometric sensitivity and the ability to discern small differences in radiated energy
Hyperspectral imaging sensors provide a highly detailed electromagnetic spectrum of agricultural fields, making it a useful tool for detecting smaller and more localized variations in important soil attributes and degradation, as well as changes in crop health and fitness. The increasing use of sensors in Agriculture 4.0 and the addition of higher resolution sensors such as hyperspectral imaging is driving the use of big data and raising concerns related to data security, data integrity, and privacy. Addressing those concerns is a major emphasis for the EU’s ATLAS program.
Agricultural interoperability and analysis system
The EU-funded Agricultural Interoperability and Analysis System (ATLAS) project aims to develop an open platform to support innovation and Agriculture 4.0. ATLAS is one of the EU’s Horizon 2020 research and innovation programs. The Fraunhofer Society is managing the project. It addresses the current lack of data interoperability in agriculture by combining agricultural equipment with sensor systems and data analysis. The resulting platform is expected to support the integration of hardware and software interoperability from a wide variety of sensor systems and amplify the benefits of digital agriculture. ATLAS aims to develop an open interoperability network for agricultural applications and build a sustainable ecosystem for innovative data-driven agriculture (Figure 3). ATLAS is building on networks of in-the-field sensors and multi-sensor systems to provide the big data needed to realize Agriculture 4.0.
The ATLAS platform is expected to support the flexible combination of agricultural machinery, sensor systems, and data analysis tools to overcome the current lack of interoperability and enable farmers to increase their productivity sustainably by using the most advanced digital technology and data. ATLAS will also define layers of hardware and software to allow the acquisition and sharing of data from a multitude of sensors and the analysis of this data using a variety of dedicated analysis approaches. The program will demonstrate the benefits of Agriculture 4.0 through a series of pilot studies across the agricultural value chain and end by defining the next generation of standards needed to continue the growing adoption of data-driven architecture.
Wireless sensor networks and security
Wireless Sensor Networks (WSNs) and Internet of Things (IoT) solutions are extensively used in Agriculture 4.0, providing numerous benefits to the farmers. However, the interconnection among diverse sensors and network devices, which can contain unpatched or outdated firmware or software, creates opportunities for network insecurities and opens various attack vectors, including device attacks, data attacks, privacy attacks, network attacks, and so on.
The increasing use of automation and even autonomous operations to improve yields also raises safety concerns. In addition to ATLAS, the European Union’s Horizon 2020 research and innovation programs focus on developing network traffic monitoring and classification tools for use in Agriculture 4.0 systems. Effective traffic monitoring is expected to play an essential role in protecting assistants and users from the impacts of network attacks. Network traffic analysis and classification tools are being developed for Agriculture 4.0 based on Machine Learning (ML) methodologies to help mitigate the threats to WSNs and other IoT-connected assets.
Summary
The deployment of Agriculture 4.0 relies on the increasing use of WSNs to improve yields and reduce costs for farmers. It also requires the development of new sensor modalities, such as plant wearables using graphene-based and FBG sensors, and the expansion of existing sensor modalities, such as the move from multispectral to hyperspectral imaging. The EU’s ATLAS program is designed to improve interoperability and realize the maximum benefit from the growing diversity of sensor and data analysis technologies. Improvements in network security will also be essential to ensure data security, integrity, and privacy in Agriculture 4.0.
References:
5 Ways to Use Multispectral Imagery in Agriculture, Coptrz
Advances in hyperspectral sensing in agriculture, Special Agriculture 4.0
Agriculture 4.0 and Smart Sensors. The Scientific Evolution of Digital Agriculture: Challenges and Opportunities, MDPI sensors
ATLAS, Agricultural Interoperability and Analysis System, ATLAS
Engineers make wearable sensors for plants, enabling measurements of water use in crops, Iowa State University
Plant Wearable Sensors Based on FBG Technology for Growth and Microclimate Monitoring, MDPI sensors
Precision Agriculture Technologies and Factors Affecting Their Adoption, US Department of Agriculture
