The growing use of devices connected to agricultural equipment and computer systems has brought us the possibility of real-time management of the most diverse agricultural operations, such as precision agriculture, irrigation, mechanization, fuel supply, meteorology, among other applications, which are common today. One of the most successful applications among those mentioned above is the management of mechanized agricultural operations through telemetry.

The cost of agricultural operations has increased considerably in recent times due to the increase in the value of fuel and also the cost of acquiring agricultural machinery, so the management of the machinery fleet is increasingly important in the planning, execution and control of agricultural operations. , as it directly involves the profitability of agricultural production.

Telemetry systems collect, through sensors, information regarding the performance of machines, such as engine speed and temperature; fuel consumption and equipment location, allowing several performance variables to be calculated automatically, such as idle time, stopping points, efficiency during operation, area worked, among others. However, there are some difficulties related to the connectivity of these systems due to the low coverage of the cellular network, an obstacle that has been overcome by some providers through the use of investment in private local networks (3G, 4G, LoraWan, Zigbee et cetera).

In this way, every telemetry project for operations must include an in-depth study of the objectives that the organization intends to make when carrying out such an investment. When introducing operations management systems, we must start with a detailed study of agricultural processes and the variables involved in each of the processes, in order to enable the creation of analysis metrics and relevant indicators.

These metrics and key performance indicators must be continuously monitored by the operations control team and make it possible to quickly detect, in the case of efficient monitoring, deviations in some processes, improving quality and reducing operating costs. According to the Oxford Dictionary, key performance indicator (KPI) is a quantifiable measure used to assess the success of an organization, employee, et cetera in meeting performance objectives. Key performance indicators are important metrics that quantify the company's performance, based on its organizational objectives, comparing the results obtained with those planned.

They must be embedded in automated systems that make it possible to capture and share information in an agile way, monitoring the results in real time, and make decisions with more security, making the necessary adjustments, developing opportunities and measuring their progress that reflects a certain period. Thus, the proper definition of the calculation period is of fundamental importance. The definition of key indicators must follow well-defined criteria so that they are representative and in adequate quantity for efficient management, so the following issues must be taken into account when choosing indicators:

What do we really need to measure?

What will be measured?

What data is available to the team?

What is the interval between measurements?

How often should we inform managers?

What is considered success?

How to identify if performance is trending away from objectives?

What data would stakeholders be most interested in?

The general concepts of performance indicators include the Index, which is the number that represents the performance obtained in a process by the performance indicators, the objective, which are the values to be targeted in a predetermined period, using the performance indicators and the acceptable tolerance limit for the variation of values, outside of which they indicate that the conduct of the process is critical. As a result, some actions must be taken.

Having defined the key indicators, the next step is to make the team that carries out the monitoring and control extract knowledge from the data transformation, that is, collect a set of information (data), make it public through some form of communication (information) and then understanding the fact for decision making (knowledge).

The correct use of the data collected in abundance in real time in the field gives us the possibility of performing predictive analysis of operations. Predictive analytics is the use of technologies to anticipate future behaviors or estimate unknown results, based on current and historical data, based on statistical models, machine learning, data mining and artificial intelligence.

Based on these analyses, we no longer make decisions based solely on intuition, managing to establish a more solid prognosis for each action, considering a greater amount of data from different sources, creating different scenarios and evaluations. One of the first applications is related to the pattern of failures in machines and equipment, allowing a previous action on the equipment, reducing lost time and maintenance costs.

Despite the benefits already possible for telemetry systems available on the market, there are some barriers to be overcome. The first of these concerns the hiring of manpower, a key element in the proper use of installed computing resources. The lack of integration of the different systems is another problem to be solved, since the information is dispersed in several places, making it difficult to correctly visualize how they interact in the operation.

However, the kick-off of this new world, the analytical world, in the management of agricultural operations has already been given, and there are companies reaping good results from well-structured projects, with qualified and dedicated personnel and adequate tools, managing to increase efficiency and sustainability. operation and in order to reduce operating costs and consumption of fossil fuels.