What drones will herd sheep and how computers determine the ripeness of fruits

Overview of streaming technologies used in agriculture.

By 2050, the world’s population will reach 9.6 billion people. In order for food to be in abundance, food production must increase by 70%. One of the key trends that will help in this is streaming agriculture.

Streaming Services transmit data from one device to another seamlessly streaming. The data can be different: from reports on user actions to telemetry and indicators of IoT devices.

In agriculture, drones, IoT and computer vision are used for streaming, while artificial intelligence helps to process the data.

IoT sensors

Market size of IoT devices is about $5.6 billion. And, according to expert estimates, the implementation of IoT solutions will allow agricultural industry to grow 20% faster.

According to PwC, IoT can reduce livestock deaths by 15%. For example, sensors on collars and ear tags are able to monitor biorhythms (even predict the start of the reproductive cycle) and the activity of each animal (in order to change food at the right time). With their help, you can identify sick individuals at an early stage of infection in order to quickly separate them from the herd and thus prevent an epidemic.

Thanks to the Internet of Things, biologists study genomes and the microclimate, thereby improving the quality of the crop. IoT use to regulate temperature, humidity, lighting and soil conditions in greenhouses. For example, the robots of the Dutch company HortiKey scan tomato bushes, measuring size, ripeness and number of fruits.

How to increase egg production using temperature and air sensors, Denis Muravyov, founder and CEO of the Russian startup GoodWAN, spoke in the release of the Digital Garden podcast. The entrepreneur explains in detail in which farms it is profitable to use IoT gadgets, where to buy them and for how much – so as not to overpay. Listen to episode you can here.


An example of field mapping by drone, www.vc.ru
An example of field mapping by drone, www.vc.ru

Them use for mapping and surveying crops. For example, drones with infrared cameras look for diseased plants in the pictures. The latter reflect less infrared radiation and look like green areas on the spectrum. Thanks to this, farmers can use less pesticides, that is, apply them pointwise – only on affected plants.

Analysis of images from satellites and drones allows predict pest attacks, identify patterns in their activity. The AI ​​analyzes the signs of an insect attack by learning from images of past attacks. Such system used, for example, in India to protect cotton fields. She classifies and counts pests in photographs of insect traps. And in some Asian countries in a similar way analyze palm gardens.

AT RSHB DIGITAL podcast “Digital Garden”, the founder and CEO of the startup Agrofly, Sergey Terekhin, told whether it is profitable for a farmer in Russia to buy a drone today or is it better to rent it. And also – how you can use flying robots – including for personal purposes.

More drones create field maps, watch shoots, irrigate plantations. For example, Swiss drones Sense Fly, which use multispectral analysis to assess crop health, has enabled French farmers to increase their yields by 10%. And the agricultural drone eBee allowed the team of the Timiryazev State Agrarian University to reduce the use of nitrogen fertilizers by 20%.

Drones looking for and counting sick or injured animals, find those who have strayed from the herd and even help to graze the cattle. A striking example is the technology of autonomous drones developed by an Israeli company Be Free Agro.

In the same place, in Israel, a startup Tevel develops drones for picking fruit. The robot picks ripe fruit, sorts and puts them into boxes that are ready for sale. Such drones – fruit pickers can work both day and night.

A program called JOE (she got her name from a shepherd dog) moves up to 1000 head of cattle over large areas with just one drone. JOE determines the obstacles in the terrain and the location of the animals, then building the best route for them. The use of drones cuts up to half of all company costs.

According to Global Market Insights, the global agricultural drone market will exceed $1 billion by 2024.

computer vision

By 2026 spending on AI in agriculture grow up four times, mainly due to the introduction of computer vision technologies (the ability of cameras not only to record a video sequence, but also to determine what exactly is happening on it).

For example, the application algorithm wineview familiar with images of “good” bunches of grapes and “bad” ones and detects them: he identifies the latter by twisted leaves, yellowness and red spots. And the cameras of the Brazilian company Cromai “sort” fruit on images by color, shape and texture, determining their ripeness. it reduces shipping losses: if you harvest a little earlier, the fruits will reach the store fresher, among them there will be less overripe fruits.

AI analyzes vineyard leaves.  In this way, the disease of twisted leaves of the vine, yellowness and red spots can be detected.  www.vc.ru
AI analyzes vineyard leaves. In this way, the disease of twisted leaves of the vine, yellowness and red spots can be detected. www.vc.ru

The computer eye is capable of measure the actual amount of water in a plantso that the farmer can water it only when necessary, save himself and ultimately reduce the cost of the final product for the buyer. smart cameras also find weeds and harmful insectsso that farmers can use chemicals locally, reducing their number, according to some estimates, up to 90%. If the soil needs fertilizer, the system will let you know and even recommend the right one to the farmer.

Alfiya Kayumova, co-founder of the agrotech startup Green Growth RSHB DIGITAL podcast explained what precision farming is and what role computer vision plays in it. Using the example of her company, Alfiya also told how to attract investments in an agrotech start-up and what traps to expect on the way of its development.

Artificial intelligence

AI can determine the weight of chickens, pigs and cattle to predict the best slaughter time and reduce fattening costs. In addition, computer vision evaluates the characteristics of milking and helps increase milk yield.

So, in Tatarstan, the company smart farm develops a dairy farm management system based on computer vision. With the help of artificial intelligence, she analyzes the behavior of animals around the clock based on camera recordings – determines the optimal time for feeding, signs of estrus, and warns farm workers about deviations.

The hour is short when AI will be able to control the full cycle of production, and a person will not even need to control its work – only to get results. In 2022, the organizers of the international competition Autonomous Greenhouse Challenge gave the participants a difficult task – to come up with a system in which the machine can grow lettuce on its own (control watering, feed plants, regulate the gas composition in the greenhouse, etc.). Everything should be so autonomous that a person I didn’t even have to press the buttons.

In the final of the competition, the Russian team of Rosselkhozbank and MIPT took second place — they managed to grow a “stand-alone lettuce” in a greenhouse and get a crop with the highest revenue. The participants themselves vividly spoke about this experiment in RSHB DIGITAL podcast.

Estimated Agribusiness Revenue from AI by 2030 will increase 18 times – up to 11.2 billion dollars.

Similar Posts

Leave a Reply