Dynamic Decision Support Tool Research Articles

Hybrid fuzzy-Bayesian decision support tool for dynamic project scheduling and control under uncertainty

Construction projects are complex and highly uncertain due to their special characteristics. Considering the incapability of current project scheduling methods in addressing the associated uncertainty, there is an emerging need for a dynamic scheduling and monitoring tool during project implementation. This paper addresses this need by proposing a dynamic decision support tool which employs fuzzy set theory to assess the combinatory effect of multiple risk factors on activities duration and Bayesian network to control and predict productivity under uncertainty. Advantages of proposed decision support tool comparing to existing approaches are: (1) calculating the fuzzy weighted average through an exact analytical solution which avoids creating incorrect fuzzy membership functions; (2) employing Type-2 interval-valued fuzzy numbers to model the uncertainty in expert evaluations, and (3) utilizing the advantages of Bayesian network in updating prior probability distributions of duration and predict productivity with actual activity data. Proposed model contributes to knowledge that could help minimize the schedule overrun and improve overall project performance. To elaborate the methodology, modelling and updating procedures of a group of activities in a simulated bridge project is presented as a case study.

Maize yield and nitrate loss prediction with machine learning algorithms

Pre-growing season prediction of crop production outcomes such as grain yields and nitrogen (N) losses can provide insights to farmers and agronomists to make decisions. Simulation crop models can assist in scenario planning, but their use is limited because of data requirements and long runtimes. Thus, there is a need for more computationally expedient approaches to scale up predictions. We evaluated the potential of four machine learning (ML) algorithms (LASSO Regression, Ridge Regression, random forests, Extreme Gradient Boosting, and their ensembles) as meta-models for a cropping systems simulator (APSIM) to inform future decision support tool development. We asked: (1) How well do ML meta-models predict maize yield and N losses using pre-season information? (2) How many data are needed to train ML algorithms to achieve acceptable predictions? (3) Which input data variables are most important for accurate prediction? And (4) do ensembles of ML meta-models improve prediction? The simulated dataset included more than three million data including genotype, environment and management scenarios. XGBoost was the most accurate ML model in predicting yields with a relative mean square error (RRMSE) of 13.5%, and Random forests most accurately predicted N loss at planting time, with a RRMSE of 54%. ML meta-models reasonably reproduced simulated maize yields using the information available at planting, but not N loss. They also differed in their sensitivities to the size of the training dataset. Across all ML models, yield prediction error decreased by 10%–40% as the training dataset increased from 0.5 to 1.8 million data points, whereas N loss prediction error showed no consistent pattern. ML models also differed in their sensitivities to input variables (weather, soil properties, management, initial conditions), thus depending on the data availability researchers may use a different ML model. Modest prediction improvements resulted from ML ensembles. These results can help accelerate progress in coupling simulation models and ML toward developing dynamic decision support tools for pre-season management.

Effectiveness of Interventions Aimed at Increasing Statin-Prescribing Rates in Primary Cardiovascular Disease Prevention

Statins are a cornerstone medication in cardiovascular disease prevention, but their use in clinical practice remains suboptimal, with less than half of people who are indicated for statins actually taking the medication. To perform a systematic review and synthesis of the literature on patient-oriented and physician-oriented interventions aimed at increasing statin-prescribing rates in adults without a history of cardiovascular disease. PubMed, Embase, and the Cochrane Library were searched for randomized clinical trials published between January 2000 and May 2019. Data abstraction was performed using the Cochrane Public Health Review Group's data collection template, and a narrative synthesis of study results was conducted. The risk of bias in each study was qualitatively assessed, and a funnel plot was created to further evaluate the risk of publication bias. Among 7948 citations and 128 full-text articles reviewed, 20 studies (of 109 807 patients) were included in the review. Eight trials reported a statistically significant increases in statin-prescribing rates. Among the effective trials, absolute effect sizes ranged from 4.2% (95% CI, 2.2%-6.4%) to 23% (95% CI, 7.3%-38.9%) and odds ratios from 1.29 (95% CI, 1.01-1.66) to 11.8 (95% CI, 8.8-15.9). Patient-education initiatives were the most commonly effective intervention, with 4 of 7 trials indicating increases in statin-prescribing rates. Two trials combined electronic decision-support tools with audit-and-feedback systems, both of which were effective overall. Physician-education programs without dynamic input regarding patient risk or updated treatment recommendations were generally found to be less effective. While heterogeneous in their interventions and outcomes, a number of interventions have demonstrated increases in statin-prescribing rates, with patient-education initiatives demonstrating more promising results than those focused on physician education alone. As opposed to more education about generic recommendations, tailored patient-focused and physician-focused interventions were more effective when they provided personalized cardiovascular risk information, dynamic decision-support tools, or audit-and-feedback reports in a multicomponent program. There are a number of modestly successful approaches to implement increases in rates of statin prescribing, a proven yet underused cardiovascular disease prevention class of therapy.

Environmental performance of energy systems of residential buildings: Toward sustainable communities

The evaluation of the environmental performance of energy systems used in residential buildings by applying the principles of the Life Cycle Analysis is an established methodological approach. Applying it in practice presents, however, significant interest, as a lack of available data has to be overcome. The research results presented in this paper include the analysis of the production, disposal and transportation of the materials used for the manufacturing processes of the building's energy systems, which include an oil and a gas fired boiler, split unit air conditioners, mono-Si and poly-Si PV arrays, flat plate and evacuated tube solar thermal collectors and their auxiliaries. The data needed for the analysis were taken from audits in the industries producing those systems, from related studies already published and from publicly available databases, when no other source was available. In this way, a comprehensive and fully adjustable database of the systems’ environmental impact has been created. This database can be a part of an integrated dynamic decision support tool, or it can be used in combination with tools commercially available. It can therefore assist prospective users in the selection of the appropriate energy systems that will lead to the minimization of the total environmental impact of new and existing buildings. The results are applied to a representative residential building and its systems are evaluated and analyzed for several scenarios.

Innovative contracting and the need for dynamic adaptive standards in the asset management of transportation infrastructures

High expectations are placed in the new asset management paradigm. Nevertheless, asset management should not be seen as a 'solution' but rather a tool that holds promise to improve decision-making through the proper identification of trade-offs. This paper aims at illustrating some of the key trade-offs faced by owners of transportation networks that combined with the shift towards innovative contracts make the need of dynamic adaptive standards urgent. The new contracting paradigm means a shift from directly exerting control over works of private agents, to indirectly managing them through a set of contract incentives. Innovative contracts are often seen as resulting in lower flexibility for public clients and limited possibilities for changes in standards during their duration. The use of dynamic decision-support tools for the setting of performance standards enforced in contracts could solve this dilemma. There is evidence to suggest that system dynamic models are suitable for the handling of long-term (dynamic) contracts.

Fed‐batch and perfusion culture processes: Economic, environmental, and operational feasibility under uncertainty

This article evaluates the current and future potential of batch and continuous cell culture technologies via a case study based on the commercial manufacture of monoclonal antibodies. The case study compares fed-batch culture to two perfusion technologies: spin-filter perfusion and an emerging perfusion technology utilizing alternating tangential flow (ATF) perfusion. The operational, economic, and environmental feasibility of whole bioprocesses based on these systems was evaluated using a prototype dynamic decision-support tool built at UCL encompassing process economics, discrete-event simulation and uncertainty analysis, and combined with a multi-attribute decision-making technique so as to enable a holistic assessment. The strategies were compared across a range of scales and titres so as to visualize how their ranking changes in different industry scenarios. The deterministic analysis indicated that the ATF perfusion strategy has the potential to offer cost of goods savings of 20% when compared to conventional fed-batch manufacturing processes when a fivefold increase in maximum viable cell densities was assumed. Savings were also seen when the ATF cell density dropped to a threefold increase over the fed-batch strategy for most combinations of titres and production scales. In contrast, the fed-batch strategy performed better in terms of environmental sustainability with a lower water and consumable usage profile. The impact of uncertainty and failure rates on the feasibility of the strategies was explored using Monte Carlo simulation. The risk analysis results demonstrated the enhanced robustness of the fed-batch process but also highlighted that the ATF process was still the most cost-effective option even under uncertainty. The multi-attribute decision-making analysis provided insight into the limited use of spin-filter perfusion strategies in industry. The resulting sensitivity spider plots enabled identification of the critical ratio of weightings of economic and operational benefits that affect the choice between ATF perfusion and fed-batch strategies.

The flight perturbation problem

Airlines spend considerable time, effort and financial resources on planning. It is essential to create a competitive timetable and construct a fleet and a crew schedule that utilizes these resources to the maximum. Unfortunately, it is all too common that an airline is faced with the necessity of reconstructing their schedules due to some unforeseen event, for example an aircraft breakdown or a crew member that is indisposed. In this paper, an application that can help airlines solve the complex problem of reconstructing aircraft schedules is presented. A mixed integer multicommodity flow model with side constraints is developed and further reformulated into a set packing model using the Dantzig—Wolfe decomposition. Cancellations, delays and aircraft swaps are used to resolve the perturbation, and the model ensures that the schedule returns to normal within a certain time. Two column generation schemes for heuristically solving the model are tested on real problem data obtained from a Swedish domestic airline. The computational tests show that the application is capable of presenting high quality solutions in a few seconds and therefore can be used as a dynamic decision support tool by the airlines.