Alley cropping is a specific agroforestry system, which is regarded as sustainable land use management, that could play a crucial role in climate change adaptation and mitigation. Despite its appealing attributes, farmers’ up-take of the system is slow in temperate regions. This study aims to contribute to scaling-up agroforestry through a case study in Hungary and to help to design productive alley cropping systems. We investigated which tree planting pattern of black locust (Robinia pseudoacacia L.) results in the most productive alley cropping system when intercropped with triticale (x Triticosecale W.) by statistically analysing the yields of the intercrop and of the trees in nine different layouts and by calculating land equivalent ratios (LER). There was significant difference between the treatments both in triticale and black locust yields. The more trees planted on a hectare, the higher the volume of the stand, and the less yield of triticale was observed, although the latter correlation was weak and in some cases the triticale was more productive between the trees compared with sole crop control. Eight out of nine treatments had favourable LER (0.94–1.35) when the trees were five years old. Black locust and triticale seem to be a good combination for productive alley cropping systems.

PLC (Programmable Logic Controllers) based data collection is integral to industrial automation and data acquisition processes. The pipeline – between the device and the database – is a complex system with many components and often causes some time delay, due to the synchronization and the applied hardware and software components. This delay can also lead to data loss on the database site. In this study, we aimed to examine this problem connected to the synchronized behavior of four counter networks within the TIA software for PLCs, particularly focusing on the implications of a 1 Hz (Hz) clock frequency on counter synchronization. Meticulous experiments were conducted and the integration of Node-RED, as an instrumental tool in industrial automation, facilitated counter-behavior monitoring. Recorded values of counter and timestamps were meticulously stored in four separate databases (MSSQL, MySQL, MongoDB, and Apache Cassandra) for comprehensive data analysis. Throughout the experiments, inconsistencies in the counter values were encountered, leading to the discovery of missing values that varied across the ten tests. To detect the reason, a unique delay calculation method was developed. According to our results, we were able to do post-synchronization with millisecond-level accuracy. It can help reveal the missing values that depend on the Node-RED and the PLC cycle time differences.

Ensuring the reliability and temporal accuracy of real-time data transmission in industrial systems presents significant challenges. This study evaluates the performance of a Siemens Programmable Logic Controller (PLC) transmitting data to a MongoDB database via Node-RED over 24 h and 14-day intervals. Key issues observed include time drift, timestamp misalignment, and forward/backward time jumps, mainly resulting from Node-RED’s internal timing adjustments. These anomalies compromised the integrity of time-sensitive data. A significant disruption on day 8 due to a power outage introduced data gaps and required manual system recovery. Additional spikes in missing data were observed after day 12. The Predictive Missing Value (PMV) model addressed these gaps. The model achieved strong accuracy at larger intervals (e.g., 5 min) but showed reduced performance at finer resolutions (1–2 min) due to the irregularity of data patterns. This research highlights the difficulty of maintaining temporal consistency in long-term, real-time systems. It also evaluates the PMV model’s effectiveness in mitigating data loss while acknowledging its limitations under complex timing disruptions.

Industrial monitoring pipelines that log programmable logic controller (PLC) states to a database for machine learning inherit the noise introduced by the sampling process itself, yet noisy-label learning research is almost always evaluated against synthetically corrupted labels rather than real sampling noise. This study evaluates a physically derived noise-rate predictor, P_edge, by training decision tree and logistic regression classifiers on real, non-synthetic noisy labels from a Festo Modular Production System–Process Automation (MPS PA) industrial automation testbed—a four-station laboratory production line used for Industry 4.0 research and teaching—and assessing them against independently logged true PLC states across four stations. A controlled temporal-subsampling experiment, which varies the effective observation interval while holding the classification task fixed, shows that F1 score (the harmonic mean of precision and recall, used here in place of raw accuracy because of class imbalance in the target signals) against ground truth declines substantially for one process (0.951 to a sweep minimum of 0.679, 0.723 at the highest sampling condition tested), replicated across two target signals and both classifiers, while two other processes remain robust (F1 at or above 0.94), consistent with strong correlated features (correlations at or above 0.98), and one target signal shows no learnable baseline (F1 at or below 0.652) regardless of noise. These results indicate that a physically derived sampling-risk index can help anticipate machine learning sensitivity in industrial automation, but only when the target task is both learnable and lacks a redundant shortcut feature.