Smart Farming

Innovations in Smart Farming with 5G

5G, Sensor Technology and Robotics Digitize Agriculture

Smart Farming

Animal welfare, environmental protection, sustainability, rising costs and a shortage of employees – the demands in agriculture are high. Fertilizers and pesticides are to be reduced by up to 90% or avoided completely, farm animals are to be better monitored, and fruits and vegetables are to be grown sustainably. At the same time, the risk of crop losses due to extreme weather events, water shortages and pest infestations is increasing.

Projects by the Fraunhofer IFF, the Eastern Switzerland University of Applied Sciences and the Bavarian State Research Center for Agriculture demonstrate solutions for agricultural businesses.

The prerequisites are the powerful mobile communications standard 5G, sensor technology and robotics.

5G, Sensor Technology and Robotics Digitize Agriculture

Technologies for Smart Farming

Whether in the field, in livestock farming or in the use of agricultural machinery – since the introduction of the fifth generation of mobile communications, new perspectives have opened up for farmers in many areas. This is because high-performance mobile communications can be used to simultaneously equip all the devices required for the Internet of Things (IoT), such as sensors, tracking systems, drones, computers or smartphones. In combination with the cloud, Big Data, artificial intelligence, high-resolution image recognition, robotics, satellite technology and GPS, farming can become smarter and thus more sustainable and efficient.

The basic prerequisite for machine-to-machine (M2M) communication in smart farming is sufficient communication capability of the autonomous machines and vehicles used on the basis of 5G. This enables large volumes of data to be transmitted in real time and almost without delay.

Digital Networking Safeguards Agricultural Production

Even the smallest deviations in the production process can lead to major quality losses in food production. This applies in particular to food products that have to be processed immediately – such as milk. Therefore, holistic approaches for the permanent analysis of all critical values are necessary. The solution is the digital networking of devices, machines and systems.

For example, the farmer receives important data on soil conditions, the health of dairy cattle, weather forecasts or the degree of plant maturity, which reliably help in risk assessment. At the same time, processes can be simplified in this way and made more efficient and easier to plan.

Three pilot projects and implemented applications illustrate the potential and diverse applications of new technologies in the field and in the barn. In particular, the automation of regularly repeated work processes is at the center of the measures for digitization.

Wireless IoT technologies and the 5G mobile communications standard are enabling new solutions for agricultural production.

M2M communication of autonomous machines and robots requires the fast transmission of large amounts of data in real time.

Targeted Weed Control

Efficient Weed Control Thanks to 5G Connectivity

Outlook: Environmentally Friendly Crop Protection

The bitter dock reaches a growth height of between 50 and 120 centimeters and displaces useful plants.

In the Swiss research project "Improving Swiss Agricultural Production through Digitization," weeds are located and destroyed using 5G-enabled drones.

Pesticide Reduction

Stump-leaved dock, also called bitter dock, is a weed. With its large leaves and deep taproots, it crowds out neighboring crops. It produces up to 60,000 seeds that remain germinable for up to 20 years. Not even cows can help with its elimination: Like all dock plants, bitter dock contains oxalic acid. For this reason, the animals do not eat it. Cereal cultivation is particularly affected. Farmers' yields are shrinking. Until now, the weed has been removed laboriously by hand or by the extensive use of crop protection products. This is set to change in the future.

Researchers at the Eastern Switzerland University of Applied Sciences (OST) are working together with fenaco, Agroscope, Sunrise UPC and Huawei on an Innosuisse project for targeted control of bitter dock using drones and agricultural robots.

Efficient Weed Control Thanks to 5G Connectivity

The meadow is photographed in high resolution with a drone and the image data is transmitted to a cloud via a 5G network. The data transfer is made possible by a built-in radio module. Artificial intelligence software analyzes the images and creates a high-resolution map from the data.

Using machine learning, the program recognizes bitter dock at different stages of growth and calculates their coordinates with centimeter precision. The data is transmitted to a robot in the field. The robot navigates to the weeds in the field via GPS. Each individual plant is removed efficiently and in an environmentally friendly manner via spot spraying.

Outlook: Environmentally Friendly Crop Protection

According to OST researchers, the use of robotics can reduce herbicide use by up to 90 percent. However, this is not enough for the scientists: Another goal of this project is to eliminate the use of pesticides altogether. Hot water could serve as an alternative. This would be particularly interesting for organic farms, but could also be used by conventional farmers.

Further trials with field robots will have to show how effective hot water is in the fight against bitter dock. One thing is certain: bitter dock is just the beginning. The goal is a complete inventory of all plants in a meadow. This would allow the solution to be extended to other weeds in the future.

Plant Monitoring

The Fraunhofer IFF has developed the solution HawkSpex, which records plant health by means of spectral analyses. Costly laboratory analyses are no longer necessary.

Spectral analysis identifies diseased or undernourished plants. Targeted fertilization or the application of plant protection products is possible.

Plant Monitoring with Soft Sensors

The Fraunhofer Institute for Factory Operation and Automation IFF is researching modern sensor solutions for agriculture. The goal is to increase ecological sustainability and efficiency. The result is HawkSpex, an adaptive smart sensor technology. This records plant condition without laboratory analysis. The technology was tested in viticulture in the "BigGrape" project.

At the heart of the solution are soft sensors, or virtual sensors. These are created by combining hardware sensors and software- based evaluations with machine learning methods. They detect not only the light visible to the human eye, but also infrared and ultraviolet radiation.

Intelligent Spectral Analysis Captures Plant Health

Spectral analysis captures the light spectrum emitted by crops. A "spectral fingerprint" is created. Based on this, a mathematical algorithm calculates the material composition of the plants. Depending on their state of health, they form different substances, which become visible with the spectral analysis. Other parameters, such as an undersupply of water or a pest infestation, are also detected thanks to the non-invasive measurement of plant metabolism.

Farms Increase Yields with Sensor Technology

Artificial intelligence-based data analysis helps agricultural businesses increase plant health and productivity. The AI can distinguish between healthy and diseased plants. The farmer can treat, irrigate or fertilize affected plants in a targeted manner. Another advantage of HawkSpex: the intelligent soft sensor solution is flexibly configurable and fits into all agricultural equipment and systems, such as harvesters, drones or small aircrafts. Existing sensors and their data can also be integrated into the system.

Health Data Monitoring

Detecting Sick Farm Animals Early with Technology

Herd management is an enormous challenge for many farmers. The productivity of livestock and the economic success in milk production depend on the health of the animals. Often, farmers detect sick animals too late. Digital applications can support the collection, monitoring and analysis of health data and warn livestock farmers at an early stage.

Livestock Monitoring in Real Time with Sensor Technology

The starting point is sensor systems fitted in the ear, on the neck or on the foot (pedometer) of the animals. These systems continuously measure parameters such as resting time, standing time, number of steps, chewing time, feeding time, body temperature and number of drinking cycles. The data is transmitted in real time to the farmer's mobile phone or computer and analyzed by artificial intelligence assistance systems.

In the event of irregularities, the system immediately sounds the alarm: for example, longer lying times and reduced locomotor activity can be an early indication of illness. A drop in body temperature, reduction in ruminant activity or changes in the animal's locomotor activity are signs of impending calving. In addition, there are special heat detection systems that recognize the typical behavior of cows in heat.

Digital Herd Management Optimized with AI

Farmers also have the option of not only storing their data on the farm computer on-site, but also using inter-farm networked systems. This allows third parties such as external breeding organizations or consultants to access the data – provided they have the appropriate authorization.

Artificial intelligence can be used to evaluate the ever-growing data networks and draw conclusions about changes in individual animals and the entire livestock population. Feeding problems or reduced fertility can also be detected at an early stage. Artificial intelligence thus makes an important contribution to improving digital herd management.

In the "DigiMilch" project of the Bavarian State Research Center for Agriculture, sensors capture the health status of animals. Signs of illness trigger warnings to farmers.

Digital Herd Management Optimized with AI

Farm animals are equipped with sensors that monitor animal health.

The Future of Smart Farming

In the future, the harvesting process, for example, could be fully automated. The farmer is then responsible for monitoring the processes and quality control in the field.

Innovative smart farming applications like those of the Innosuisse project are testing the future use of smart farming machines.

Further Smart Farming Applications in Practice

The multitude of available new technologies are already making the vision of smart farming a reality on some farms: Combine harvesters autonomously order the transporter during harvest and transmit the location where the crop is to be delivered. Feeding and milking robots take over tasks in the barn. During the journey, they collect important data, for example on the amount of feed, and transmit it to the farmer's computer or smartphone.

Some farmers also fertilize their fields with the help of GPS systems. The system locates the tractor's position and prevents fertilizer from being applied outside the field or from being applied twice.

Outlook: 5G Applications in Agriculture

Digital technologies based on 5G are already being used in some agricultural businesses.

Pilot projects from research and development show: In the future, further application scenarios in agriculture are conceivable – for example, smart irrigation systems that detect soil moisture or robots that identify weeds in the field and eliminate them in a predictive, precise and cost-saving manner. Harvesting could also be fully automated in the future. In this way, smart farming benefits consumers, the environment, animals and, last but not least, the farmers themselves.

However, the digital revolution in agriculture has only just begun. Many steps are still needed to reach the goal – a fully digitized agriculture. The top priority is development in other areas, for example, to create the legal conditions for autonomous driving of agricultural machinery. In these cases, it is unclear who is liable in the event of an accident.