Machine learning (ML) has revolutionized resource management in agriculture by analyzing vast amounts of data and creating precise predictive models. Precision agriculture improves agricultural productivity and profitability while reducing costs and environmental impact. However, ML implementation faces challenges such as managing large volumes of data and adequate infrastructure. Despite significant advances in ML applications in sustainable agriculture, there is still a lack of deep and systematic understanding in several areas. Challenges include integrating data sources and adapting models to local conditions. This research aims to identify research trends and key players associated with ML use in sustainable agriculture. A systematic review was conducted using the PRISMA methodology by a bibliometric analysis to capture relevant studies from the Scopus and Web of Science databases. The study analyzed the ML literature in sustainable agriculture between 2007 and 2025, identifying 124 articles that meet the criteria for certainty assessment. The findings show a quadratic polynomial growth in the publication of articles on ML in sustainable agriculture, with a notable increase of up to 91% per year. The most productive years were 2024, 2022, and 2023, demonstrating a growing interest in the field. The study highlights the importance of integrating data from multiple sources for improved decision making, soil health monitoring, and understanding the interaction between climate, topography, and soil properties with agricultural land use and crop patterns. Furthermore, ML in sustainable agriculture has evolved from understanding weather data to integrating advanced technologies like the Internet of Things, remote sensing, and smart farming. Finally, the research agenda highlights the need for the deepening and expansion of predominant concepts, such as deep learning and smart farming, to develop more detailed and specialized studies and explore new applications to maximize the benefits of ML in agricultural sustainability.

Measuring climatic conditions is a fundamental task for a wide array of scientific and practical fields. Weather variables change depending on position and time, especially in tropical zones without seasons. Additionally, the increasing development of precision or smart agriculture makes it necessary to improve the measurement systems while widely distributing them at the location of crops. For these reasons, in this work, the design, construction and fabrication of an adaptable autonomous solar-powered climatic station with wireless 3G or WiFi communication is presented. The station measures relative humidity, temperature, atmospheric pressure, precipitation, wind speed, and light radiation. In addition, the system monitors the charge state of the main battery and the energy generated by the photovoltaic module to act as a reference cell for solar energy generation capability and agrivoltaic potential in the installation area. The station can be remotely controlled and reconfigured. The collected data from all sensors can be uploaded to the cloud in real-time. This initiative aims at enhancing the development of free and open source hardware that can be used by the agricultural sector and that allows professionals in the area to improve harvest yield and production conditions.

The measurement of outdoor environmental and climatic variables is needed for many applications such as precision agriculture, environmental pollution monitoring, and the study of ecosystems. Some sensors deployed for these purposes such as temperature, relative humidity, atmospheric pressure, and carbon dioxide sensors require protection from climate factors to avoid bias. Radiation shields hold and protect sensors to avoid this bias, but commercial systems are limited, often expensive, and difficult to implement in low-cost contexts or large deployments for collaborative sensing. To overcome these challenges, this work presents an open source, easily adapted and customized design of a radiation shield. The device can be fabricated with inexpensive off-the-shelf parts and 3-D printed components and can be adapted to protect and isolate different types of sensors. Two material approaches are tested here: polylactic acid (PLA), the most common 3-D printing filament, and acrylonitrile styrene acrylate (ASA), which is known to offer better resistance against UV radiation, greater hardness, and generally higher resistance to degradation. To validate the designs, the two prototypes were installed on a custom outdoor meteorological system and temperature and humidity measurements were made in several locations for one month and compared against a proprietary system and a system with no shield. The 3-D printed materials were also both tested multiple times for one month for UV stability of their mechanical properties, their optical transmission and deformation under outdoor high-heat conditions. The results showed that ASA is the preferred material for this design and that the open source radiation shield could match the performance of proprietary systems. The open source system can be constructed for about nine US dollars, which enables mass development of flexible weather stations for monitoring needed in smart agriculture.

Unmanned Aerial Vehicles (UAVs), also known as drones, have had an exponential evolution in recent times due in large part to the development of technologies that enhance the development of these devices. This has resulted in increasingly affordable and better-equipped artifacts, which implies their application in new fields such as agriculture, transport, monitoring, and aerial photography. However, drones have also been used in terrorist acts, privacy violations, and espionage, in addition to involuntary accidents in high-risk zones such as airports. In response to these events, multiple technologies have been introduced to control and monitor the airspace in order to ensure protection in risk areas. This paper is a review of the state of the art of the techniques, methods, and algorithms used in video, radiofrequency, and audio-based applications to detect UAVs and Unmanned Aircraft Systems (UAS). This study can serve as a starting point to develop future drone detection systems with the most convenient technologies that meet certain requirements of optimal scalability, portability, reliability, and availability.

In lighting engineering, it has become necessary to develop new acquisition devices aligned with today’s technological needs, so that they can be easily integrated with IoT (Internet of things) systems. This allows recording lighting data in real-time either to perform analysis, estimate lighting quality measurements, or interact with intelligent lighting systems. Here a portable low-cost IoT hyperspectral acquisition device is presented, which can be used to perform measurements both indoors and outdoors. Since the device is part of IoT technologies, it has wireless communication and allows raw data to be sent to the cloud or to perform calculations directly within the device to be transmitted later. In addition, information on the hardware general design is provided and the device measurements uniformity is validated using the ASTM G-173-03 solar spectra. The proposed device can be used outdoors since its container is IP67 and it can simultaneously measure incident and diffuse radiation. On the other hand, the recorded data can be used for light source spectrum recovery, and to calculate light quality measurements. The hardware and software are customizable, which is an advantage over commercially available devices. The final cost of the device is 271.95 USD.

Color and lighting quality assessment allow users to reliably examine the quality of light sources and the changes in the lighting conditions. Technological advance has led to an increase in the development of intelligent systems that are permanently connected to the Internet and that allow access to real time-data. Smart lighting systems have become a subject of great interest given their potential in different application areas, such as optimization for object reflectance and damage reduction to artwork. It has become necessary to develop new sensors that are aligned with the technological needs, that are easily integrated into IoT (Internet of Things) or smart lightning systems, and that allow estimating lighting quality measurements in real time. Here, a low-cost IoT multi-spectral acquisition device with wireless communication is presented. Insights into the overall system design and a validation of its operation is provided to allow reproducibility and independent testing of the proposed device. The device can be used for light source spectrum recovery and estimation of light quality and color rendering measures. The total cost of the device is below 130 USD, and customizability provides a great advantage over commercially available devices.