How new remote sensing systems and solutions can assist in day-to-day sugarcane farming

One of the biggest concerns of the sugarcane producer is the productivity of the plantation. It determines whether the investments have paid off and will generate profits and whether the products will have the desired quality and quantity.

The productivity issue is special when analyzing the costs of producing sugarcane, sugar, and ethanol. According to a survey carried out annually by the Institute of Research and Continuing Education in Economics and Management (Pecege), the expenses for a sugarcane plant's agricultural and industrial parts are pretty high and depend on several factors.

In 2019/20, for example, an expense of R $ 102.16 per ton of cane was found, which is equivalent to R $ 7,870.20 per hectare planted. Even if this value is below that observed in the previous harvest, it means an essential investment for producers, who, in general, are looking for different ways to reduce it. And the reduction in relative costs can be made with cost cuts — potentially harmful for the future of the cane field — or with greater productivity.

“This is the permanent challenge for plants and independent producers”, says the senior business consultant for the Swiss company Gamaya in Brazil, Silvio Santos. Not dissociated from this is the objective of companies to generate profit, after all, “they are companies”. “Productivity represents more revenue and profit has to be combined with reduced expenses,” he says.

But the question is: how to increase productivity and reduce expenses? The possibilities available today in the sector are diverse. In addition to the choice of newer or even genetically modified varieties, which eliminate pests, for example, the positive effect of digital technologies in all stages of sugarcane management has proved to be an increasingly crucial ally in the search for increased productivity and reduced costs.

“Digital technologies are especially efficient in controlling the use of chemicals,” says Santos.

Specializing in using advanced technologies such as remote sensing, artificial intelligence, and machine learning to develop remote monitoring systems for agriculture, Gamaya recently launched Canefit, a set of solutions aimed at the cane field in Brazil. The company proposes to make life easier for the producer, which may impact expenses with sugarcane production.

Accurate and scalable handling

When analyzing several sugarcane fields across the country in the past five years, Gamaya researchers found that the plants need proactive solutions, hampered by the size of Brazilian plantations. Santos explains that the distortion or variability of the information must be in a small range to achieve satisfactory levels of accuracy.

Still taking into account the area of ​​cane fields, he says that every solution needs to be generated on a large scale and the decision-making must be quick. “Talking to the sugarcane mills, we understood that they need this: accurate solutions, on a large scale and with speed,” he sums up.

Besides, the Gamaya team noted the need for streamlining operational processes and convenience within organizations. “If at the push of a button or with few activities you can have the solution, you have a lot of conveniences, you have more time to analyze things, and the decision is much more assertive,” he reports.

Canefit products were developed based on these principles. For the time being, five solutions are offered that start from the advanced analysis of images produced by drones or satellites, in different formats, focused on the main challenges of each phase of the culture, from planting to harvest: Failures and basic planting lines; Detection of weeds by drone; Monitoring of weeds by satellite; Sugarcane monitoring; and Precision harvest lines.

Santos explains that one of the main differentials of Gamaya’s solutions is remote sensing, AI and agronomic modeling.

The Swiss company pioneered developing a portable hyperspectral camera, which can be attached to drones and other vehicles. The technology goes beyond the traditional digital (RGB) and multispectral cameras, which generally make up the satellites, allowing the visualization of a greater spectrum of bands and, with this, a significant level of detail — which enables the detection of several problems of the cane fields.

While conventional digital cameras “see” 3 bands and multispectral cameras, from 8 to 12, hyperspectral cameras capture up to 40 bands.

Gamaya solutions do not depend on the use of hyperspectral cameras. The same technology was used to develop the products and, combined with exclusive algorithms created to automate the image analysis processes, allowed its adaptation to the reality of the market, which uses equipment with other standards.

“Thus, our image processing systems are now able to identify, with much greater detail, variations that may indicate anomalies in the plantations, even in photos taken with conventional equipment (RGB) and satellites,” he explains.

Besides, Santos reinforces that all technology must be convenient. For this reason, the company developed the Gamaya Viewer platform, in which the users of its products interact and view the information obtained “in a straightforward and very accessible way.” According to the consultant, the objective is that even a lay producer understands the information required to manage the cane field better.

“If you have a product that is to identify the weed, the person, in a very playful way, can see where it is, what the statistics are, the types of herbs in the different areas of a plant, be it small, medium or large. You don’t need a lot of technical knowledge to understand the platform ”, explains Santos.

The Gamaya Viewer can be accessed online and offline by computers and mobile devices.

Reduction or optimization of input costs

The example given by Gamaya Brasil’s senior business consultant may even refer to two of the Canefit solutions: Drone weed detection and Satellite weed monitoring.

As the name says, the first is a system for the detection and mapping of infestation of sugarcane crops by weeds, with which “accurate information is generated for planning and executing pest management operations, resulting in more rational use and sustainable use of pesticides.” It is recommended for use primarily in the early stages of crop growth, between 40 and 120 days.

Santos explains that weed control is a suitable parameter for measuring the influence of these solutions on plant costs. “In the past, when there was no such solution, the sugarcane plant did not know where the weeds were, so it used the herbicide in the whole plot because there were some in the middle that someone, looking at, found,” he reports.

Thus, he explains, there was an excessive and unnecessary expenditure on herbicides. “With our solution, the producer discovers that he has 10% of the area with weeds, and our product informs him, in a georeferenced way, where each of them is”, he states and adds: “In this way, using equipment for precision application, like the drone, but it can be a tractor or a tricycle, they [the producers] put the herbicide in the sprayer, which applies it exactly on top of that weed.”

This can save more than 70% in spending on herbicides, thanks to both the precision and the incredible speed to find these problems. “It is good for the environment and the plant’s pocket. It is good for everyone because the producer only uses the herbicide where he needs it,” he concludes.

The relevance of this product lies in the fact that the control of weeds represents one of the highest agricultural costs in the cane fields. According to the most recent study by Pecege, spending on pesticide applications — the most efficient way to combat invaders — reaches 21.8% of the total cost of inputs at the plants.

According to Gamaya, poor control can mean significant losses, with loss of productivity and compromised sugarcane longevity. According to the Brazilian Agricultural Research Corporation (Embrapa), weeds cause problems that raise the cost of producing sugarcane by up to 30%.

Also, in this area, Gamaya has the option of satellite weed monitoring. “This solution offers a constant and comprehensive look, issuing reports with periodic updating (monthly, in the first version of the product) of the entire property, in which it is possible to compare the levels of infestation by plots and to define, quickly, the spraying strategies at each stage,” describes the company.

According to the company, combining both technologies increases the level of detail of crop observation. While the satellites bring an enlarged view of the set, the drones fly in a targeted manner, at the points indicated by the first, to confirm the local infestation's location and level.

According to the company, satellite monitoring is intended for the moment considered most critical for combat action, when the return on investment in spraying brings more returns, preventing a future loss of productivity. It is indicated in the stages of crop development, up to five or six months after planting or cutting.

Santos explains that this solution (as well as all of Gamaya’s) is constantly evolving through machine learning: “Once the product is launched, it is always in automatic development.” The proposal for the coming months of the company is to launch more and more solutions, following the plants' returns.

Still, this year, the new solutions launched should include, in addition to detection, the classification of weeds by category, between broad or narrow leaves, and by species, which determines the use of the best herbicide.

Close, real-time monitoring

The satellite, with multispectral images, is also the technology used in the sugarcane monitoring solution. The producer can monitor the growth of the plants and receive weekly reports indicating anomalies and eventual problems facing them before more severe losses can occur.

The images, analyzed by the company’s algorithm, identify which factors may be causing a problem in the field, be they biotic or abiotic, diseases, erosion, lack of nutrients in the soil, or any other difficulty.

The unevenness identifies the problem as a whole of the crop. Thus, the plant can locate it and send a team to check the specified point, optimizing the work in the field. The solution is recommended for all stages of sugarcane development, allowing continuous monitoring of all its areas.

In an upcoming version, the product must include an alarm system. “Every time we identify an anomaly, a problem in the cane field, we will send a message informing the people designated by the plant,” he explains.

In Santos’s opinion, this product can help many managers solve problems faster and on a much larger detection scale. “When the cane grows a little, no one else goes in the middle to see the problem. You only find out when you will harvest, and there you have already had the economic damage ”, he details.

Another solution that works with image analysis, but with drones, is the Harvest lines with precision. In this case, the algorithms are used to find the best paths for agricultural machinery and generate accurate data for autopilots.

The technology makes it possible to identify the actual planting lines in the cane fields and export maps compatible with the equipment of sprayers and harvesters, which need to follow smooth trajectories and avoid trampling the plants left in the area. “That line that has been stepped on will not produce cane in the next cycle, so the producer loses productivity with this,” observes Santos.

These maps can also be used in fertilization and spraying operations, optimizing handling throughout the production cycle and increasing the efficiency and quality of the harvest.

Agility in the correction of the failure

Like precision harvest lines, the Faults and primary planting lines solution also uses aerial image processing made with drones to generate information that helps the producer plan and execute more efficient replanting operations. According to Gamaya, the objective is to provide better profitability in the field, with an increase in the productivity of each field and greater longevity in the cane field.

“The traditional method of identifying failures is done by sending teams to the field, who do the work manually and by sampling,” reports the company, pointing out that this process can be time-consuming and costly, mainly due to the more significant hand requirement work.

With the solution recommended for the initial stages of the crop, the Gamaya platform diagnoses the length of the planting lines, identifying and measuring the operational failures precisely and with the georeferenced location of the detected problems.

Sílvio Santos explains that the program identifies, via image, a space in the cane field where there is no cane, although culture should be present. Once the absence of the plant has been identified, the plant has two possibilities for action.

The first consists of replanting to cover the areas without the cane. “This means that, if it does not cover, that fault will remain there for five years, and it will be five years without producing cane in that space,” comments the executive.

On the other hand, the second option is to use the fault information and its location to feed the fertilizer application sprayer, making it avoid that area, resulting in an economy of inputs.

“The producer wins in both ways. Either save by not applying or increase productivity by replanting. Thus, this product thinks of two facets of gain for the plant, calculated based on an accurate diagnosis ”, concludes Santos.

Analyzes carried out by the Gamaya team based on the use of the product Failures and primary planting lines indicate that, in a plot with 20% of failures, a replanting operation on time could generate a profit of around R $ 2,700 hectare each year. Considering a five-year sugarcane cycle, the profit would surpass the mark of R $ 13,500 per hectare.

Suppose the decision is not to carry out the planting. In that case, the data collected by Gamaya from one of its customers — a large plant in the state of São Paulo — indicates savings of more than 10% in the use of fertilizers since the product is no longer applied in areas where it was noted that there were no plants.

The savings with Canefit products, according to Santos, are more felt in the costs of inputs or transportation, as it avoids unnecessary expenses. “Besides, it is not only the detection of problems but a more assertive application,” he says.

He adds: “Every time the producer has an accurate diagnosis of a problem, he designs the solution in an economical and in a great amount of time, so our competence is to help the plants to identify, very quickly and very accurately, the problems they need to decide to solve.”

Brazilian cane as a study

Canefit products were initially developed at Gamaya’s headquarters in Switzerland. “Once the technical concept, computer science, image processing was calibrated, validating several hypotheses of the hyperspectral image of our specific cameras for this, we went to the field here in Brazil to design products for sugarcane.”

According to Santos, the company chose the country because it is one of the leading agricultural countries in the world, and sugarcane, as is mainly concentrated in some nearby states. “We were able to have satellite images of an entire state, such as São Paulo, more adjacent, and this gives a critical mass to make the products have competitive costs for the user,” he reports.

In Brazil, the products were developed with the producers over almost five years. Gamaya contacted the plant technicians, who said precisely the specific problems with that unit and identified those most familiar to several plants.

“We have developed a relationship of understanding with the producers. Although we are a Swiss company, we were wearing a boot for a long time, walking in the cane fields and talking to the plant technicians ”, Sílvio Santos (Gamaya Brasil)

During this period, the company defined the criteria for solutions as the north, which should encompass different moments in the development of sugarcane; rapid diagnostics, in a matter of one week at most; and scale applicability, with analysis and reproducibility in huge areas.

Santos reinforces that Brazilian sugarcane fields are significant, and it is impossible to make a diagnosis with the naked eye. “Even on a motorcycle, car, donkey, or helicopter, it is tough for the human eye to capture what we get through hyperspectral cameras, or with drones, with RGB cameras, or even with satellite,” he adds.

Thus, he believes that the demand for solutions like this will increase more and more, thanks to the peculiarities of the cane that hinder efficient and cheaper management of the plantation.

“In a universe of 350 plants, there should not be 50 at all, using these technologies in full. But adoption is very fast. You need these initial technologies so that others can enter in a second moment ”, Sílvio Santos (Gamaya Brasil)

“These technologies are coming in to help, from the point of view of adding a great economic value for the plants and their managers,” defends the consultant, who adds: “There is a phrase that says we can see far away with our cameras,” and with our ability to process data in a customized way for the interests of the sugarcane mills and producers, we can design customized solutions for this world.”

Thus, he believes that there are still “many other possibilities” that the company is researching to deliver more sophisticated products in crop monitoring.