Exploring Digital Innovations in Agriculture: A Pathway to Sustainable Food Production and Resource Management

Today’s agriculture faces numerous challenges due to climate change, a growing population and the need to increase food productivity. In response to these issues, digital innovation is becoming an essential tool that can significantly improve the efficiency of the agricultural sector. The digitization of agriculture, which includes technologies such as the Internet of Things (IoT), machine learning, drones and advanced monitoring systems, offers new opportunities that can help improve production processes, resource management and environmental protection.

This SI titled “Digital Innovations in Agriculture-Series II” presents a collection of studies that illustrate the diversity of applications of innovative technologies in agriculture. The articles in this Special Issue explore such areas as the accurate monitoring of plant and animal health, the optimization of production processes and the development of intelligent management systems. Through the use of modern analytical and diagnostic tools, these studies show how innovation can lead to a more informed and efficient use of resources [1,2].

The opportunities created by these technologies not only increase production efficiency, but also contribute to biodiversity conservation and ecosystem sustainability. In addition, data analysis methods and artificial intelligence-based approaches enable better forecasting and risk management, which is critical in the face of a changing climate and increasing extreme events. This Special Issue represents an important step toward understanding the impact of digital technologies on the future of agriculture and agri-food sustainability [3,4,5].

We hope that the research collected in it will inspire both researchers and industry practitioners to further explore the potential of innovative solutions in agriculture and their practical application in daily operations.

The perishable nature of commodities in the agri-food supply chain poses challenges for producers and marketers. Using an ecological routing model for the distribution of fresh agricultural products using an adaptive hybrid nutcracker optimization algorithm (AH-NOA) allows for increased population diversity. Thanks to k-means clustering, the algorithm effectively overcomes local optima, leading to significant reductions in costs and CO₂ emissions. Studies have indicated that AH-NOA significantly improves global search efficiency, making this method promising. In addition, MushR’s gourmet mushroom breeding automation system achieved 91.7% accuracy in identifying ready-to-harvest mushrooms using Raspberry Pi-based harvesting mechanics. This approach shows how technologies can revolutionize the harvesting process in food production [6,7].

Detecting oil palm pests using UAVs provides a modern solution for agriculture. The use of drones along with a multispectral camera allows for the quick and accurate analysis of infestation rates. Studies have shown a 100% F1 score in classifying healthy and low levels of infestation, demonstrating the effectiveness of this method. Rapid response through the use of UAVs can save time and resources compared to traditional detection methods. This type of technology is key in crop management, increasing the efficiency of palm oil monitoring. The results show the importance of introducing innovative solutions in agricultural production [8].

The models used to forecast potato yields in Poland showed significant potential by analyzing an integrated data set from 36 fields. Three models were developed: non-satellite, satellite and hybrid, and the best result emerged from the hybrid model, achieving a MAPE of 5.85%. Advanced data analysis techniques, such as PCA and outlier detection, helped improve forecasting performance, and the hybrid model combining data from all sources yielded the best results. This approach to yield analysis can significantly improve forecasting accuracy in agriculture. The results of this research suggest that machine learning plays a key role in improving crop management. Ultimately, innovative yield forecasting tools have the potential to optimize agricultural production [9].

A detection system for recognizing blueberry fruit maturity, based on the improved YOLOv5 algorithm, achieved impressive results, such as an average recall of 92.0% and an average precision of 91.5%. The use of the ShuffleNet module and Convolutional Blocks of Attention (CBAM) contributed to better feature fusion, and the system’s speed is 67.1 frames per second. The method, with its lower computational requirements, is more suitable for use in edge devices. A comparison with other algorithms, such as SSD and Faster R-CNN, showed better results with a lighter model weight. This method has potential for applications in intelligent fruit detection systems in orchards, which can support modern agriculture. These achievements show how technology can revolutionize fruit picking [10].

Evaluating the differential revision responses of multiple wheat genotypes to water stress reveals significant differences in tolerance and plant growth. Studies that used polyethylene glycol at different concentrations applied machine learning techniques to effectively predict key growth parameters. The best results in predicting water absorption and germination rates were obtained using the elastic-net model. In addition, an analysis of the performance of classification techniques showed that a Gaussian process classifier was best for estimating root length. The results of this research are crucial for wheat genetics and breeding in the face of a changing climate. The application of new analytical techniques indicates the possibility of crop improvement under harsh conditions [11].

The use of computer vision technology to classify apple varieties has yielded exceptional results, achieving an accuracy as high as 97.48% using the DenseNet201 model. The implementation of CNN architectures and transfer learning in fruit classification confirmed the effectiveness of this method. Further analysis of the deep features extracted from the model allowed the use of traditional classifiers such as SVM and MLP, which contributed to an even higher accuracy of 99.77%. This study also considered the impact of dimensionality reduction on classification performance. These achievements underscore the importance of artificial intelligence in improving classification quality in an agricultural product. The findings point to the need for further research in apple variety diversity, which could enhance analytical capabilities [12].

Wheat Teacher is a wheat ear detector that uses semi-supervised methods and employs pseudo-tagging and consistency regularization. Experiments conducted on a set of GWHD2021 showed that Wheat Teacher achieved a mAP0.5 score of 92.8%, using only 20% of the labeled data. Such a result outperformed the results of two full-surveillance-based detection models that were trained on 100% of the labeled data, with a difference of no more than 1%. In addition, Wheat Teacher improved mAP0.5 by 2.1%, achieving 37.7% across different proportions of labeled data usage. This research underscores Wheat Teacher’s significant potential in effective wheat ear detection, which can contribute to better breeding strategies. The system indicates the potential for applications of semi-supervised technology in the context of increasing efficiency in wheat crop management [13].

The introduction of a mobile pig weight prediction system, based on Raspberry Pi and 3D camera data analysis, made it possible to automate the weighing process. The system achieved an RMSE of about 10.702, confirming its high precision. The use of 3D technology contributes to reducing labor costs in animal husbandry. With effective weight measurement, this study illustrates the innovation of measurement methods in animal production. The results of this research are applicable to increasing productivity in animal husbandry. Such systems can significantly improve production efficiency in the animal industry [14].

In addition, drone technologies in insecticide application in young rice orchards show the importance of precision agriculture. The introduction of a prescription map allows chemicals to be concentrated in areas with the highest loss, reducing their use at the same time. Studies have shown that the use of drones increases the efficiency of crop protection, leading to a lower percentage of rice damage compared to control areas. This approach contributes to the development of modern crop protection methods. The results of this research are a step toward more sustainable agricultural practices that can benefit the environment. The use of drones in crop protection opens up new opportunities for more efficient agricultural production [15].

Studies on radio signal propagation in sensor networks have shown that different models do not respond adequately to signal attenuation in dense vegetation. Comparative analyses with actual measurements in a tapioca field in Colombia have shown significant differences in network design. Machine learning techniques, including random forests and K-NN, significantly improved the prediction of signal attenuation. The final findings may be useful in planning new wireless networks in agriculture. This study is a step toward improving communication technologies in the agricultural sector. The development of such technologies has a major impact on the efficiency of agricultural resource management [16].

The use of a neuromorphic computer to estimate the weight and length of fish in smart aquaponic systems has demonstrated the ability to classify more than 84 million samples in just one second. The SpikoPoniC system, based on this research, has a 3369-fold acceleration compared to traditional computers. This enables the intelligent monitoring of aquaponics, which is crucial for future commercial solutions. This study highlights the importance of hardware optimization in aquaculture applications. Thus, it becomes clear how modern technologies can revolutionize the monitoring of aquatic resources. Innovative approaches to fish monitoring have great potential in aquaponics systems [17].

Faced with declining populations of Apis mellifera bees, which are crucial for pollination, the authors of [18] present an IoT system for beekeepers in developing countries. The use of low-cost devices to monitor hive parameters, such as temperature and humidity, allows the early detection of problems. The efficiency of the system is enhanced by processing data using an advanced machine learning model. This study also takes into account the analysis of energy consumption and network traffic. Ultimately, the proposed solution improves the accuracy of mite detection. Conclusions address the role of technology in pollinator protection and food security [18].

A system for monitoring the feed and water intake of sows provides information on their health and performance in real time. The use of open technologies and an electronic feeder consisting of a data processing program based on Spark and Flink brings significant benefits. The system allows 20,000 sows to be monitored simultaneously, which improves the management of their breeding. Studies have confirmed that with optimization, the system achieves a TPS of 6399 pcs/s for 10,000 sows. The results indicate the technology’s potential in intensive agriculture. Such solutions can significantly contribute to improving efficiency in livestock production [19].

Today’s agriculture is facing challenges from a growing population and climate change, requiring increased efficiency in food production. The digitization of agriculture, through the use of technologies such as IoT, drones and machine learning, offers solutions to increase the efficiency and sustainability of the agribusiness sector. The research included in “Digital Innovations in Agriculture-Series II” demonstrates the wide range of applications of innovative techniques that can lead to a more sustainable use of resources and better crop and livestock management. Modern analytical and predictive technologies play a key role in protecting the environment and increasing production efficiency, which is crucial in the context of a changing climate. The development of these technologies indicates the significant impact of digitalization on the future of agriculture and the potential in optimizing production processes.