Real-time Decision Support System Research Articles

Machine vision and artificial intelligence for plant growth stress detection and monitoring: A review

The agricultural sector faces increasing challenges in ensuring food security and optimizing crop yield, necessitating innovative solutions for early detection and mitigation of plant growth stress. The integration of advanced imaging technologies with artificial intelligence (AI) has emerged as a powerful tool for non-invasive, real-time monitoring of plant health. The objective of this paper was to review the application of machine vision and AI in identifying and classifying plant growth stress, with a focus on stressors, datasets, and the use of intelligent algorithms. The significance of plant growth stress induced by environmental variables, including temperature, light, nutrient deficiencies, and water supply were addressed and the conventional stress detection methodologies, underscores their inherent limitations, and establishes the groundwork for the exploration of state-of-the-art technologies in stress assessment. Various sensor technologies were explored, encompassing traditional RGB cameras, multispectral and hyperspectral sensors, and thermal imaging, each capable of capturing distinct stress signatures. Machine vision, leveraging high-resolution imaging and spectroscopy, offers detailed insights into plant physiological responses. Coupled with AI approaches such as deep learning, neural networks, and pattern recognition, machine vision enables the automated analysis of vast datasets, enhancing the accuracy and speed of stress detection. The recent advancements in image processing techniques tailored for plant stress identification were focused and discussed the role of feature extraction, classification, and predictive modelling in achieving robust results. The potentials of AI in plant stress physiology and its role in overcoming the limitations of traditional methods, and the use of unsupervised identification of visual symptoms to quantify stress severity, allowing for the identification of different types of plant stress were studied. Moreover, the potentials of machine vision technology and AI for real-time monitoring and decision support systems in precision agriculture were discussed. The findings of this review would contribute to the growing field of agricultural technology, offering insights into the development of automated tools that could aid farmers and researchers in mitigating the impact of abiotic stressors on crop/plant health and productivity.

Real-time decision support for human–machine interaction in digital railway control rooms

This study proposes a real-time Decision Support System (DSS) using machine learning to enhance proactive management of Human–Machine Interaction (HMI) in safety–critical digital control rooms. The DSS provides explainable predictions and recommendations regarding near-future automation usage, customized for the railway control room management, who supervise the operations of traffic controllers (TCs). In this setting, TCs decide on the spot whether to manually or automatically open signals to regulate railway traffic, a critical aspect of ensuring punctuality and safety. This time-setting specific HMI differs across TCs and is not yet supported by a data-driven tool. The proposed DSS includes agreement levels for predictions among different modeling paradigms: linear models, tree-based models, and deep neural networks. SHAP (SHapley Additive exPlanations) values are deployed to assess the agreement level in explainability between these different modeling paradigms. The prescriptions are based on the HMI of well-performing peers. We implement the DSS as proof of concept at the Belgian railway infrastructure company and report end-user feedback on the perception, the operational impact, and the inclusion of agreement levels.

The Transformative Impact of Big Data in Healthcare: Improving Outcomes, Safety, and Efficiencies

The integration of Big Data analytics into the healthcare sector has initiated a period of unparalleled transformation, significantly influencing patient care, research, and healthcare administration. This detailed synthesis, derived from a systematic literature review of 31 articles published between 2013 and 2023, examines the extensive impact of Big Data within the healthcare industry. It addresses how Big Data enhances personalized medicine and optimizes treatment protocols, promotes the use of predictive analytics for the early detection and management of diseases, and bolsters patient safety through the implementation of real-time monitoring and decision support systems. The review also confronts the challenges and limitations associated with Big Data's application, including concerns over data privacy and security, ethical dilemmas, interoperability issues, and the necessity for both skilled personnel and advanced technological infrastructure. Furthermore, this overview ventures into future prospects and technological advancements, emphasizing the pivotal role of emerging technologies such as artificial intelligence (AI), machine learning, and blockchain in heralding a new era of healthcare. It highlights Big Data's instrumental role in supporting public health measures and preparing for pandemics, alongside forecasting the ongoing influence of Big Data in fostering innovative healthcare paradigms. This analysis advocates for a concerted effort among healthcare professionals, policymakers, and technology experts to fully leverage Big Data's capabilities in healthcare, aiming to enhance health outcomes, increase operational efficiencies, and ensure the long-term viability of healthcare systems.

TCCC Decision Support With Machine Learning Prediction ofHemorrhage Risk, Shock Probability.

Expected future delays in evacuation during near-peer conflicts in remote locales are expected to require extended care including prolonged field care over hours to days. Such delays can increase potential complications, such as insufficient blood flow (shock), bloodstream infection (sepsis), internal bleeding (hemorrhage), and require more complex treatment beyond stabilization. The Trauma Triage Treatment and Training Decision Support (4TDS) system is a real-time decision support system to monitor casualty health and identify such complications. The 4TDS software prototype operates on an Android smart phone or tablet configured for use in the DoD Nett Warrior program. It includes machine learning models to evaluate trends in six vital signs streamed from a sensor placed on a casualty to identify shock probability, internal hemorrhage risk, and need for a massive transfusion. The project team used a mixed methods approach to create and evaluate the system including literature review, rapid prototyping, design requirements review, agile development, an algorithm "silent test," and usability assessments with novice to expert medics from all three services. Both models, shock (showing an accuracy of 0.83) and hemorrhage/massive transfusion protocol, were successfully validated using externally collected data. All usability assessment participants completed refresher training scenarios and were able to accurately assess a simulated casualty's condition using the phone prototype. Mean responses to statements on evaluation criteria [e.g., fit with Tactical Combat Casualty Care (TCCC), ease of use, and decision confidence] fell at five or above on a 7-point scale, indicating strong support. Participatory design ensured 4TDS and machine learning models reflect medic and clinician mental models and work processes and built support among potential users should the system transition to operational use. Validation results can support 4TDS readiness for FDA 510k clearance as a Class II medical device.

Improving McDiarmid Tree Performance for Predicting Heart Disease from Data Streams with Missing and Meaningless Values

Building machine learning algorithms for real-time clinical decision support systems has become a current research hotspot. The success of these algorithms depends on their ability to handle data stream characteristics. For example, we cite as characteristics the large amounts of data, the high speed and rate of incoming data, and the change in data nature and distribution over time. The Very Fast Decision Tree (VFDT) is a method for incrementally building decision trees. Since its proposition in the literature, it has become one of the most popular tools for data stream classification. This paper aims to optimize a new version of VFDT called McDiarmid Tree (MT) for the early prediction of heart diseases in real-time clinical decision support systems. The proposed method for improving MT performance consists of two main mechanisms: detecting the presence of missing and meaningless values in data attributes and handling the impact of this presence. The proposed MT has been compared with MT and VFDT. Simulation results show that the proposed MT attains significantly higher prediction accuracy with less time and model cost (RAM-Hours) than the other two algorithms.

Development and clinical validation of real-time artificial intelligence diagnostic companion for fetal ultrasound examination.

Prenatal diagnosis of a rare disease on ultrasound relies on a physician's ability to remember an intractable amount of knowledge. We developed a real-time decision support system (DSS) that suggests, at each step of the examination, the next phenotypic feature to assess, optimizing the diagnostic pathway to the smallest number of possible diagnoses. The objective of this study was to evaluate the performance of this real-time DSS using clinical data. This validation study was conducted on a database of 549 perinatal phenotypes collected from two referral centers (one in France and one in the UK). Inclusion criteria were: at least one anomaly was visible on fetal ultrasound after 11 weeks' gestation; the anomaly was confirmed postnatally; an associated rare disease was confirmed or ruled out based on postnatal/postmortem investigation, including physical examination, genetic testing and imaging; and, when confirmed, the syndrome was known by the DSS software. The cases were assessed retrospectively by the software, using either the full phenotype as a single input, or a stepwise input of phenotypic features, as prompted by the software, mimicking its use in a real-life clinical setting. Adjudication of discordant cases, in which there was disagreement between the DSS output and the postnatally confirmed ('ascertained') diagnosis, was performed by a panel of external experts. The proportion of ascertained diagnoses within the software's top-10 differential diagnoses output was evaluated, as well as the sensitivity and specificity of the software to select correctly as its best guess a syndromic or isolated condition. The dataset covered 110/408 (27%) diagnoses within the software's database, yielding a cumulative prevalence of 83%. For syndromic cases, the ascertained diagnosis was within the top-10 list in 93% and 83% of cases using the full-phenotype and stepwise input, respectively, after adjudication. The full-phenotype and stepwise approaches were associated, respectively, with a specificity of 94% and 96% and a sensitivity of 99% and 84%. The stepwise approach required an average of 13 queries to reach the final set of diagnoses. The DSS showed high performance when applied to real-world data. This validation study suggests that such software can improve perinatal care, efficiently providing complex and otherwise overlooked knowledge to care-providers involved in ultrasound-based prenatal diagnosis. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Developing a supervised learning-based simulation method as a decision support tool for rebalancing problems in bike-sharing systems

Bike-Sharing Systems (BSSs) have exploded in popularity worldwide because of their beneficial impacts on traffic, pollution levels, and public health, which has resulted in moving toward a green city. The rebalancing problem, as one of the most important operational problems of such systems, deals with planning bike distribution at different stations. Regarding conducted studies, simulation models are the most common tool for analyzing BSSs and decision-making. This popularity is based on simulation’s capabilities in modeling complexities of systems and uncertainty. Despite their advantages, lack of quickness is a significant drawback of simulation-based methods, making them inefficient for real-time decision-making processes, especially in large-scale and complex systems. In this regard, this paper introduces a Supervised Learning-Based Simulation (SLBS) method as an alternative to the conventional simulation-based methods dealing with rebalancing problems. SLBS is a huge step toward developing a real-time Decision Support System (DSS) for BSSs. For developing SLBS, firstly, we have developed a simulation model based on real-world assumptions of station-based BSSs and big data analysis of CitiBike, a well-known BSS located in New York City. The simulation model is able to calculate the number of unsatisfied demands (either number of failed pick-ups (FPs) or failed drop-offs (FDs)) as a result of different rebalancing plans. Then, the developed simulation model was used to generate quality and quantity training datasets to train Machine Learning (ML) algorithms involved in SLBS. While these ML models are trained once, SLBS will be capable of predicting the number of unsatisfied demands without running highly time-intensive simulations replications. The results obtained from a wide range of conducted experiments indicate that SLBS, up to 300 times faster than simulation models, can provide predictions with over 90% of R2 Score.

Flood modelling for a real-time decision support system of the covered Lower Paillons River, Nice, France

Abstract Nice Metropolis in Alpes Maritimes, France is prone to flood. The city is crossed by the Lower Paillons River (LPR). Its discharge for a return period of 100 years is estimated at 794 m3/s. Part of the river is covered by 2 km. In addition, there are two retention storages in the river bed and a floodable road tunnel on the left bank. Due to the increase in urban development, flood management is challenging. An existing decision support system, Aquavar, uses DHI Mike tools to reproduce runoff for the Lower Var River in the same region. To extend this system to the LPR and reinforce flood management, a new modelling tool adapted to the characteristics of the LPR is needed. Consequently, this research utilizes the DHI MIKEPLUS tool to develop a 1D -2D coupled model for real-time flood management. The results demonstrate that flood events like those in 2017 and 2019 were correctly reproduced. The linear regression R2 is above 0.8 for all stations. It was also estimated that the covered river should stay clean to avoid widespread flooding in the urban area. Overall, the model is useful for simulating flow in real time and can help sustain urban development.

Plant Disease Detection and Classification Using Deep Learning

Abstract: Agriculture is the backbone of every country in the world. In India, most of the rural population still depends on agriculture. The agricultural sector provides major employment in rural areas. Furthermore, it contributes a significant amount to India’s gross domestic product (GDP). Therefore, protecting and enhancing the agricultural sector helps in the development of India’s economy. In this work, a real-time decision support system integrated with a camera sensor module was designed and developed for the identification of plant disease. This research proposes an intelligent method for plant disease classification using image processing techniques. The proposed method aims to assist farmers and experts in identifying and diagnosing plant diseases efficiently and accurately. The system first obtains images of the plant leaves from different perspectives and then preprocesses the images to enhance the quality and remove noise. The preprocessed images are then subjected to feature extraction using a deep convolutional neural network (CNN) model. The features extracted from the CNN are then fed into a classifier for the classification of plant diseases. The proposed method is evaluated using a dataset of plant images with three different types of diseases. The results obtained show that the proposed method achieves high accuracy in the classification of plant diseases, making it a useful tool for plant disease diagnosis and control. The proposed method can be integrated into a mobile application or web based platform for use by farmers and botanists.

Decision-making during urgent phase of a nuclear accident under extreme conditions

Disasters when occurs result in a disruption of the normal life of society or community and have a severe impact on human life, health, environment, and social and economic activities. Ionizing radiations are a concern associated with nuclear accidents by their very nature, and managing a nuclear catastrophe adds additional pressure on decision-makers to address the immediate and long-term effects of ionizing radiations. Decision-making is more difficult during the early stages of a nuclear disaster because of the high level of uncertainty in the situation and the unavailability of detailed information. We introduce a systematic methodology named ROSYNA, a real-time decision support system for a nuclear accident to be considered for such an uncertain situation. The core of this decision support system is use of a decision table specific to various situations by assessing the decision problem attributes. The prototype applicability of the method was implemented to the Fukushima situation and a comparison was made with HERCA-WENRA and the IAEA approaches. With due consideration of uncertainties regarding the situation during the early phase of a nuclear emergency, we recommend considering the implementation of protective measures by utilizing the ROSYNA philosophy in the decision-making process. The decision problem attributes, which are the key element of this methodology need to be assessed using professional judgment based on the nature of the accident, the information that can be gleaned from the nuclear facility, radiation monitoring, prior understanding of such occurrences and pre-established response arrangements.

JRODOS customization for Pakistan and assessment of the consequences of a hypothetical nuclear accident

Since the Chernobyl disaster in 1986, decision support systems and modelling tools have been utilized in response to nuclear and radiological emergencies. The java-based real-time online decision support system (JRODOS) is a decision support tool that can be utilised in response to an emergency in managing off-site radiological consequences. This article documents the customization and use of JRODOS for Pakistan. JRODOS was tailored to the local Pakistan conditions, and a case study of a theoretical nuclear power plant accident was used to assess JRODOS’s feasibility as a decision support tool. A worst-case probabilistic accident scenario was used to identify zones and areas where urgent protective actions, early protective actions and food restriction and other response actions could be required. The areas and distances identified for the implementation of protective and response actions for such a hypothetical accident were found to be in agreement with the emergency planning zones and distances suggested by the International Atomic Energy Agency (IAEA). Additionally, the implications of meteorological and source term input parameters on predicting the radiation doses to members of the public were investigated. It was identified that the output of such tools strongly depends on the availability and accuracy of the input parameters, such as radioactive release and meteorological data. Limitations and uncertainties associated with these tools need to be considered in deciding on protective and other response actions in response to a nuclear accident. As established by the IAEA, protective and other response actions need to be applied on a graded approach, taking into account the protection strategy and uncertainties and limitations in the available information and criteria, based on the conditions at the facility and off-site.

Communication architecture for real-time decision support systems in railway nodes

Rail transport nodes are complex systems providing transport for goods and people. The dispatcher providing operational control is essential for the proper functioning of these nodes. His role mainly consists of decision-making connected with the allocation of limited resources such as tracks or personnel. These decisions are complicated because of many unexpected influences, e.g., train delays. Recognizing the importance of decision support for the dispatcher's decision-making process, we had designed a real-time decision support system with a simulation model at its core. Such a system requires a high-quality information source for its operation, that needs to be efficiently acquired, processed, and used. In this paper, therefore, we address the possibilities of obtaining up-to-date information about the railway system. Moreover, we designed a communication architecture that allows the distribution of information to the designed decision support system with a simulation model in the core.

Which models for extreme flood events in Mediterranean catchments?

Over the last years, most of the Mediterranean countries have been affected by catastrophic flood events generated by never observed before extreme rainfalls. Those events are frequently convective and stationary for several hours. The recorded intensities have frequently values exceeding 100 mm/hour and can last for several hours. Under those specific conditions as during the storm ALEX of October 2020, the runoff processes are particularly intense and can generate massive floods. Mitigation strategies should be developed and based on modeling that can support the understanding of the processes and the forecast the magnitude of the flood events and the associated impacts. The choice of the relevant hydrological modeling tools is a key element. The selected ones should be able to provide the relevant information, on time and with the relevant accuracy for the decision makers. The hydrological deterministic distributed models have reached a sufficient maturity to provide an accurate representation of the processes. The storm ALEX was generating more than 500 mm in less than 6 hours and producing a devastating flood within the Vésubie valley (France) in October 2020. The analysis of the event has demonstrated the difficulty to obtain consistent field observations and the limitations of many stochastic hydrological tools. The application of the AquaVar approach, combining several distributed deterministic models, has demonstrated its capacity to generate meaningful hydrographs and to quantify process with the catchment. The proposed solution is currently implemented within a real-time decision support system and can be used in real-time.

An Accessible Clinical Decision Support System to Curtail Anesthetic Greenhouse Gases in a Large Health Network: Implementation Study.

Inhaled anesthetics in the operating room are potent greenhouse gases and are a key contributor to carbon emissions from health care facilities. Real-time clinical decision support (CDS) systems lower anesthetic gas waste by prompting anesthesia professionals to reduce fresh gas flow (FGF) when a set threshold is exceeded. However, previous CDS systems have relied on proprietary or highly customized anesthesia information management systems, significantly reducing other institutions' accessibility to the technology and thus limiting overall environmental benefit. In 2018, a CDS system that lowers anesthetic gas waste using methods that can be easily adopted by other institutions was developed at the University of California San Francisco (UCSF). This study aims to facilitate wider uptake of our CDS system and further reduce gas waste by describing the implementation of the FGF CDS toolkit at UCSF and the subsequent implementation at other medical campuses within the University of California Health network. We developed a noninterruptive active CDS system to alert anesthesia professionals when FGF rates exceeded 0.7 L per minute for common volatile anesthetics. The implementation process at UCSF was documented and assembled into an informational toolkit to aid in the integration of the CDS system at other health care institutions. Before implementation, presentation-based education initiatives were used to disseminate information regarding the safety of low FGF use and its relationship to environmental sustainability. Our FGF CDS toolkit consisted of 4 main components for implementation: sustainability-focused education of anesthesia professionals, hardware integration of the CDS technology, software build of the CDS system, and data reporting of measured outcomes. The FGF CDS system was successfully deployed at 5 University of California Health network campuses. Four of the institutions are independent from the institution that created the CDS system. The CDS system was deployed at each facility using the FGF CDS toolkit, which describes the main components of the technology and implementation. Each campus made modifications to the CDS tool to best suit their institution, emphasizing the versatility and adoptability of the technology and implementation framework. It has previously been shown that the FGF CDS system reduces anesthetic gas waste, leading to environmental and fiscal benefits. Here, we demonstrate that the CDS system can be transferred to other medical facilities using our toolkit for implementation, making the technology and associated benefits globally accessible to advance mitigation of health care-related emissions.

Land use suitability assessment for economic development at the provincial level: The case study of Yazd Province, Iran

• The concept of land use suitability (LUS) is intended to help stakeholders make better land-use decisions on a regional scale. • GIS application with a hybrid Multi-Criteria Decision Analysis (MCDA)-Rule-Based System (RBS) in LUS is recommended. • This study suggests the real-time Spatial Decision Support System (SDSS) in LUS analysis. • SDSS should be used in places where suitable geodatabases are not available. Iran's Yazd Province has expanded its economy during the past 40 years due to requests from investors and decisions made by the Provincial Planning and Development Council. This article aims to aid the council's decision-making process by presenting an analytical method for determining whether a region is suitable, unsuitable, or conditionally suitable for generating various economic activities in the form of individual suitability maps. Using a hybrid Multi-Criteria Decision Analysis (MCDA)-Rule Based System (RBS), the method is made up of the Analytic Hierarchy Process (AHP) and semi/non-compensatory decision rules in combination with GIS spatial analysis. The findings show that the province has exceeded its environmental carrying capabilities since a significant portion of its overall land area is unsuitable for most industrial and agricultural activities. The Provincial Planning and Development Council seeks to have even fewer options than to concentrate on high-tech and service sector development along major and dense corridors while relocating other economic activity to the periphery to address issues with water scarcity, air pollution, uneven infrastructure distribution, and environmental limitations. The proposed model of this study can be utilized as a real-time Spatial Decision Support System (SDSS) for the Yazd Provincial Planning and Development Council. In sequential decision-making processes, the model can examine the effects of previous decisions and adjust its findings/conclusions as needed.

Increasing situation awareness in healthcare through real-time simulation

Research into real-time simulation applications outside of manufacturing environments has extended to sociotechnical systems such as healthcare over the past decade, where a number of published studies have demonstrated proof-of-concept models for near-future resource planning. Using real-time decision-support systems, people take decisions supported by the output of simulations. However real-time simulation frameworks abstract human intervention to an “external decision-maker,” with little regard to the complexities of underlying decision-making constructs, and how design and development decisions can impact the quality of decision-support. One such construct is situation awareness (SA), which is a precursor to decision-making. It is a dynamic state of knowledge about how a situation is unfolding; one approach to enhancing situation awareness is the provision of appropriate real-time information. We argue that design, development and implementation decisions should be focused at the interface between decision-making and decision-support. This integrative literature review proposes a SA framework integrating models of SA with a technical perspective for real-time simulation, to support an understanding of the cognitive needs of users alongside technical details during the development process. The implications for the usefulness and usability of real-time decision-support tools are discussed with application to Emergency Departments.

Internet of Things based Decision Support System for Green Logistics

This work proposes an IoT-based Real-Time Decision Support System for Perishable Products. The proposed system collects data during the transportation process and will interfere in the case of failure. Three different simulation models corresponding to different configurations and mitigation plans are built. The simulation models consider decisions such as stopping transportation and rerouting shipments to minimize losses in case of failure. The three different supply chain simulation models are implemented through a case study that considers transporting a perishable fruit in the intercontinental United States. A financial and environmental analysis is conducted to show the benefits of the proposed system.

Building an Augmented Reality Experience on Top of a Smart Pavement Management System

Pavement Management Systems (PMS) offers a systematic collection, storage, analysis, and modeling of road condition data to optimize resources across a road network. Adding artificial intelligence (AI) and augmented reality (AR) to PMS could improve their technical or visual aspects. This paper tries to identify a method to improve the understanding of the consequences of the city council’s decisions in the urban pavement management system field. This paper establishes the potential of AR. It provides future maintenance and rehabilitation (M&R) actions needed based on the recommendation of the future distress in the study area. The road cracks are discovered through technical analysis, and a CityEngine model is established based on the PMS results. Additionally, in terms of visualization, this paper’s unique feature delivers the result as an AR experience. Applying the Unity game engine and importing the built CityEngine model and the embedded textures as input empowered us to provide a dynamic product in terms of data and analysis and a real-time Decision Support System (DSS) for the final users. This paper concludes that researchers need many different modules to design and implement an efficient PMS to move toward a smart PMS. The smart city concept is meaningless without a tight collaboration between all distinctive parts of each urban infrastructure management system. Additionally, this paper attempts to provide answers for researchers and an outlook for future research, the development of the proposed method, and its application in other fields

A time-aware attention model for prediction of acute kidney injury after pediatric cardiac surgery.

Acute kidney injury (AKI) is a common complication after pediatric cardiac surgery, and the early detection of AKI may allow for timely preventive or therapeutic measures. However, current AKI prediction researches pay less attention to time information among time-series clinical data and model building strategies that meet complex clinical application scenario. This study aims to develop and validate a model for predicting postoperative AKI that operates sequentially over individual time-series clinical data. A retrospective cohort of 3386 pediatric patients extracted from PIC database was used for training, calibrating, and testing purposes. A time-aware deep learning model was developed and evaluated from 3 clinical perspectives that use different data collection windows and prediction windows to answer different AKI prediction questions encountered in clinical practice. We compared our model with existing state-of-the-art models from 3 clinical perspectives using the area under the receiver operating characteristic curve (ROC AUC) and the area under the precision-recall curve (PR AUC). Our proposed model significantly outperformed the existing state-of-the-art models with an improved average performance for any AKI prediction from the 3 evaluation perspectives. This model predicted 91% of all AKI episodes using data collected at 24h after surgery, resulting in a ROC AUC of 0.908 and a PR AUC of 0.898. On average, our model predicted 83% of all AKI episodes that occurred within the different time windows in the 3 evaluation perspectives. The calibration performance of the proposed model was substantially higher than the existing state-of-the-art models. This study showed that a deep learning model can accurately predict postoperative AKI using perioperative time-series data. It has the potential to be integrated into real-time clinical decision support systems to support postoperative care planning.

Bayesian Non-Parametric Thermal Thresholds for Helicoverpa armigera and Their Integration into a Digital Plant Protection System

The development of temperature-driven pest risk thresholds is a prerequisite for the buildup and implementation of smart plant protection solutions. However, the challenge is to convert short and abrupt phenology data with limited distributional information into ecological relevant information. In this work, we present a novel approach to analyze phenology data based on non-parametric Bayesian methods and develop degree-day (DD) risk thresholds for Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) to be used in a decision support system for dry bean (Phaseolus vulgaris L.) production. The replication of each Bayesian bootstrap generates a posterior probability for each sampling set by considering that the prior unknown distribution of pest phenology is Dirichlet distribution. We computed R = 10,000 temperature-driven pest phenology replicates, to estimate the 2.5%, 50% and 95.5% percentiles (PC) of each flight generation peak in terms of heat summations. The related DD thresholds were: 114.04 (PC 2.5%) 131.8 (PC 50%) and 150.9 (PC 95.5%) for the first, 525.8 (PC 2.5%), 551.7 (PC 50%) and 577.6 (PC 95.5%) for the second and 992.7 (PC 2.5%), 1021.5 (PC 50%) and 1050 (PC 95.5%) for the third flight, respectively. The thresholds were evaluated by estimating the posterior differences between the predicted (2021) and observed (2022) phenology metrics and are in most cases in acceptable levels. The bootstrapped Bayesian risk thresholds have the advantage to be used in modeling short and noisy data sets providing tailored pest forecast without any parametric assumptions. In a second step the above thresholds were integrated to a sub-module of a digital weather-driven real time decision support system for precise pest management for dry bean crops. The system consists of a customized cloud based telemetric meteorological network, established over the border area of the Prespa National Park in Northern Greece, and delivers real time data and pest specific forecast to the end user.