Travel time prediction of urban public transportation based on detection of single routes

Improving travel time prediction for public transit effectively enhances service reliability, optimizes travel structure, and alleviates traffic problems. Its greater time-variance and uncertainty make predictions for short travel times (≤35min) more subject to be influenced by random factors. It requires higher precision and is more complicated than long-term predictions. Effectively extracting and mining real-time, accurate, reliable, and low-cost multi-source data such as GPS, AFC, and IC can provide data support for travel time prediction. Kalman filter model has high accuracy in one-step prediction and can be used to calculate a large amount of data. This paper adopts the Kalman filter as a travel time prediction model for a single bus based on single-line detection: including the travel time prediction model of route (RTM) and the stop dwell time prediction model (DTM); the evaluation criteria and indexes of the models are given. The error analysis of the prediction results is carried out based on AVL data by case study. Results show that under the precondition of multi-source data, the public transportation prediction model can meet the accuracy requirement for travel time prediction and the prediction effect of the whole route is superior to that of the route segment between stops.

Using AI and ML to predict shipment times of therapeutics, diagnostics and vaccines in e-pharmacy supply chains during COVID-19 pandemic

PurposeThis paper aims to address the pressing problem of prediction concerning shipment times of therapeutics, diagnostics and vaccines during the ongoing COVID-19 pandemic using a novel artificial intelligence (AI) and machine learning (ML) approach.Design/methodology/approachThe present study used organic real-world therapeutic supplies data of over 3 million shipments collected during the COVID-19 pandemic through a large real-world e-pharmacy. The researchers built various ML multiclass classification models, namely, random forest (RF), extra trees (XRT), decision tree (DT), multilayer perceptron (MLP), XGBoost (XGB), CatBoost (CB), linear stochastic gradient descent (SGD) and the linear Naïve Bayes (NB) and trained them on striped datasets of (source, destination, shipper) triplets. The study stacked the base models and built stacked meta-models. Subsequently, the researchers built a model zoo with a combination of the base models and stacked meta-models trained on these striped datasets. The study used 10-fold cross-validation (CV) for performance evaluation.FindingsThe findings reveal that the turn-around-time provided by therapeutic supply logistics providers is only 62.91% accurate when compared to reality. In contrast, the solution provided in this study is up to 93.5% accurate compared to reality, resulting in up to 48.62% improvement, with a clear trend of more historic data and better performance growing each week.Research limitations/implicationsThe implication of the study has shown the efficacy of ML model zoo with a combination of base models and stacked meta-models trained on striped datasets of (source, destination and shipper) triplets for predicting the shipment times of therapeutics, diagnostics and vaccines in the e-pharmacy supply chain.Originality/valueThe novelty of the study is on the real-world e-pharmacy supply chain under post-COVID-19 lockdown conditions and has come up with a novel ML ensemble stacking based model zoo to make predictions on the shipment times of therapeutics. Through this work, it is assumed that there will be greater adoption of AI and ML techniques in shipment time prediction of therapeutics in the logistics industry in the pandemic situations.

The method of predicting the maintenance time of mechanical products based on Virtual Prototype

Aim: To predict the mechanical product maintenance time is difficult in the situation of lack of physical prototype or similar products’ statistics in stage of design Method: According to the theory of time accumulative estimation method, a product maintenance time prediction method framework based on virtual prototype was constructed, which described the prediction process. The virtual maintenance environment which contains virtual prototype, virtual human and maintenance tools was developed. The virtual human’s position and posture information during the maintenance process was obtained by implementing VBScript language. Result: Basic maintenance motions that constitute the whole maintenance process were classified into 4 categories: body movement, upper limb movement, grasp/replace and operation. Based on MODAPTS (Modular arrangement of predetermined time standard) method and virtual maintenance simulation, corresponding time prediction methods for each categories were proposed. Discussion: Take a maintenance dissassembly and assembly task of engine as an example, through the comparison between the measured actual maintenance time and predicted time of several methods, feasibility and effectiveness of proposed method are verified

A Physics-Informed Convolutional Neural Network with Custom Loss Functions for Porosity Prediction in Laser Metal Deposition

Physics-informed machine learning is emerging through vast methodologies and in various applications. This paper discovers physics-based custom loss functions as an implementable solution to additive manufacturing (AM). Specifically, laser metal deposition (LMD) is an AM process where a laser beam melts deposited powder, and the dissolved particles fuse to produce metal components. Porosity, or small cavities that form in this printed structure, is generally considered one of the most destructive defects in metal AM. Traditionally, computer tomography scans measure porosity. While this is useful for understanding the nature of pore formation and its characteristics, purely physics-driven models lack real-time prediction ability. Meanwhile, a purely deep learning approach to porosity prediction leaves valuable physics knowledge behind. In this paper, a hybrid model that uses both empirical and simulated LMD data is created to show how various physics-informed loss functions impact the accuracy, precision, and recall of a baseline deep learning model for porosity prediction. In particular, some versions of the physics-informed model can improve the precision of the baseline deep learning-only model (albeit at the expense of overall accuracy).

Study on real-time prediction model of railway passenger flow based on big data technology

In order to solve the limitation of traditional offline forecasting application scenarios, the author uses a variety of big data open source frameworks and tools to combine with railway real-time data, and proposes a real-time prediction model of railway passenger flow. The model architecture is divided into four levels from bottom to top: data source layer, data transmission layer, prediction calculation layer and application layer. The main components of the model are data flow and prediction flow. Through message queue and ETL, the data process part realizes the synchronization of offline data and real-time data; through the big data technology frameworks such as Spark, Redis and Hive and the GBDT (Gradient Boosting Tree) algorithm, the prediction process partially realizes the real-time passenger flow of the train OD section prediction. The experimental results show that the model proposed by the author has certain practicability and accuracy both in performance and prediction accuracy.

Dynamical-Empirical forecast for the Indian monsoon rainfall using the NCEP Coupled Modelling System – Application for real time monsoon forecast

The performance of the National Centre for Environmental Prediction’s (NCEP) operational coupled modeling system known as the Climate Forecast System (CFS) is evaluated for the prediction of all India summer monsoon rainfall (AISMR) during June to September (JJAS). The evaluation is based on the hindcast initialized during March, April and May with 15 ensemble members each for 25 years period from 1981 to 2005.The CFS’s hindcast climatology during JJAS of March (lag-3), April (lag-2) and May (lag-1) initial conditions show mostly an identical pattern of rainfall similar to that of observed climatology with both the rainfall maxima (over the west-coast of India and over the head Bay of Bengal region) well captured, with a signification correlation coefficient between the forecast and observed climatology over the Indian monsoon region (bounded by 50°E-110°E and 10°S-35°N) covering Indian land mass and adjoining oceanic region. Although the CFS forecast rainfall is overestimated over the Indian monsoon region, the land only rainfall amount is underestimated compared to observation. The skill of the prediction of monsoon rainfall over the Indian land mass is found to be relatively weak, although it is significant at 95% with a correlation coefficient (CC) of 0.44 with April ensembles.By using CFS predicted JJAS rainfall over the regions of significant CCs, a hybrid dynamical-empirical model is developed for the real time prediction of AISMR, whose skill is found to be much higher (CC significant above 99% level) than the raw CFS forecasts. The dynamical-empirical hybrid forecast applied on real time for 2009 and 2010 monsoons are found to be much closer to the observed AISMR. Thus, when the hybrid model is used there is a correction not only to the sign of the actual forecast as in the case of 2009 monsoon but also to its magnitude and hence can be used as a better tool for the real time prediction of AISMR.