Federated learning (FL) is a promising approach for edge/IoT-based distributed machine learning, where both privacy and bandwidth efficiency are essential. However, as time progresses, edge/IoT-based FL faces challenges such as unpredictable concept drift, leading to model performance degradation and the need for frequent retraining. To address these challenges, we propose a federated learning framework designed for heterogeneous IoT devices, capable of handling continuous data distribution changes while accounting for limited storage resources. Our framework introduces a server-side drift detection method to minimize bandwidth usage and optimize retraining times, conserving IoT device resources. We also present an efficient storage management strategy to mitigate catastrophic forgetting by selectively managing incoming data streams within device constraints. Additionally, we develop an exemplar-based online continual learning algorithm that leverages class prototypes in the deep feature space to further combat catastrophic forgetting. We evaluate our framework on image classification tasks using ImageNet and CIFAR-100 datasets across four model architectures, demonstrating significant improvements in adaptation to concept drift and long-term performance stability compared to baseline FL approaches.

The public commute is essential to all urban centers and is an efficient and environment-friendly way to travel. Transit systems must become more accessible and user-friendly. Since public transit is majorly designed statically, with very few improvements coming over time, it can get stagnated, unable to update itself with changing population trends. To better understand transportation demands and make them more usable, efficient, and demographic-focused, we propose a fast, multi-layered transit simulation that primarily focuses on public transit simulation (BTE-Sim). BTE-Sim is designed based on the population demand, existing traffic conditions, and the road networks that exist in a region. The system is versatile, with the ability to run different configurations of the existing transit routes, or inculcate any new changes that may seem necessary, or even in extreme cases, new transit network design as well. In all situations, it can compare multiple transit networks and provide evaluation metrics for them. It provides detailed data on each transit vehicle, the trips it performs, its on-time performance and other necessary factors. Its highlighting feature is the considerably low computation time it requires to perform all these tasks and provide consistently reliable results.

The ability to accurately predict public transit ridership demand benefits passengers and transit agencies. Agencies will be able to reallocate buses to handle under or over-utilized bus routes, improving resource utilization, and passengers will be able to adjust and plan their schedules to avoid overcrowded buses and maintain a certain level of comfort. However, accurately predicting occupancy is a non-trivial task. Various reasons such as heterogeneity, evolving ridership patterns, exogenous events like weather, and other stochastic variables, make the task much more challenging. With the progress of big data, transit authorities now have access to real-time passenger occupancy information for their vehicles. The amount of data generated is staggering. While there is no shortage in data, it must still be cleaned, processed, augmented, and merged before any useful information can be generated. In this paper, we propose the use and fusion of data from multiple sources, cleaned, processed, and merged together, for use in training machine learning models to predict transit ridership. We use data that spans a 2-year period (2020-2022) incorporating transit, weather, traffic, and calendar data. The resulting data, which equates to 17 million observations, is used to train separate models for the trip and stop level prediction. We evaluate our approach on real-world transit data provided by the public transit agency of Nashville, TN. We demonstrate that the trip level model based on Xgboost and the stop level model based on LSTM outperform the baseline statistical model across the entire transit service day.

Emergency response is highly dependent on the time of incident reporting. Unfortunately, the traditional approach to receiving incident reports (e.g., calling 911 in the USA) has time delays. Crowdsourcing platforms such as Waze provide an opportunity for early identification of incidents. However, detecting incidents from crowdsourced data streams is difficult due to the challenges of noise and uncertainty associated with such data. Further, simply optimizing over detection accuracy can compromise spatial-temporal localization of the inference, thereby making such approaches infeasible for real-world deployment. This paper presents a novel problem formulation and solution approach for practitioner-centered incident detection using crowdsourced data by using emergency response management as a case-study. The proposed approach CROME (Crowdsourced Multi-objective Event Detection) quantifies the relationship between the performance metrics of incident classification (e.g., F1 score) and the requirements of model practitioners (e.g., 1 km. radius for incident detection). First, we show how crowdsourced reports, ground-truth historical data, and other relevant determinants such as traffic and weather can be used together in a Convolutional Neural Network (CNN) architecture for early detection of emergency incidents. Then, we use a Pareto optimization-based approach to optimize the output of the CNN in tandem with practitioner-centric parameters to balance detection accuracy and spatial-temporal localization. Finally, we demonstrate the applicability of this approach using crowdsourced data from Waze and traffic accident reports from Nashville, TN, USA. Our experiments demonstrate that the proposed approach outperforms existing approaches in incident detection while simultaneously optimizing the needs for real-world deployment and usability.

This paper presents a data-driven approach for predicting the propagation of traffic congestion at road segments as a function of the congestion in their neighboring segments. In the past, this problem has mostly been addressed by modelling the traffic congestion over some standard physical phenomenon through which it is difficult to capture all the modalities of such a dynamic and complex system. While other recent works have focused on applying a generalized data-driven technique on the whole network at once, they often ignore intersection characteristics. On the contrary, we propose a city-wide ensemble of intersection level connected LSTM models and propose mechanisms for identifying congestion events using the predictions from the networks. To reduce the search space of likely congestion sinks we use the likelihood of congestion propagation in neighboring road segments of a congestion source that we learn from the past historical data. We validated our congestion forecasting framework on the real world traffic data of Nashville, USA and identified the onset of congestion in each of the neighboring segments of any congestion source with an average precision of 0.9269 and an average recall of 0.9118 tested over ten congestion events.

Non-recurring traffic congestion is caused by temporary disruptions, such as accidents, sports games, adverse weather, etc. We use data related to real-time traffic speed, jam factors (a traffic congestion indicator), and events collected over a year from Nashville, TN to train a multi-layered deep neural network. The traffic dataset contains over 900 million data records. The network is thereafter used to classify the real-time data and identify anomalous operations. Compared with traditional approaches of using statistical or machine learning techniques, our model reaches an accuracy of 98.73 percent when identifying traffic congestion caused by football games. Our approach first encodes the traffic across a region as a scaled image. After that the image data from different timestamps is fused with event- and time-related data. Then a crossover operator is used as a data augmentation method to generate training datasets with more balanced classes. Finally, we use the receiver operating characteristic (ROC) analysis to tune the sensitivity of the classifier. We present the analysis of the training time and the inference time separately.

Effective public transit operations are one of the fundamental requirements for a modern community. Recently, a number of transit agencies have started integrating automated vehicle locators in their fleet, which provides a real-time estimate of the time of arrival. In this paper, we use the data collected over several months from one such transit system and show how this data can be potentially used to learn long term patterns of travel time. More specifically, we study the effect of weather and other factors such as traffic on the transit system delay. These models can later be used to understand the seasonal variations and to design adaptive and transient transit schedules. Towards this goal, we also propose an online architecture called DelayRadar. The novelty of DelayRadar lies in three aspects: (1) a data store that collects and integrates real-time and static data from multiple data sources, (2) a predictive statistical model that analyzes the data to make predictions on transit travel time, and (3) a decision making framework to develop an optimal transit schedule based on variable forecasts related to traffic, weather, and other impactful factors. This paper focuses on identifying the model with the best predictive accuracy to be used in DelayRadar. According to the preliminary study results, we are able to explain more than 70% of the variance in the bus travel time and we can make future travel predictions with an out-of-sample error of 4.8 minutes with information on the bus schedule, traffic, and weather.