Indiana University-Purdue University Indianapolis (IUPUI) / On-time delivery of supplies to industry is essential because delays can disrupt production schedules. The aim of the proposed application is to predict transit times for outbound logistics thereby allowing suppliers to plan for timely mitigation of risks during shipment planning. The predictive model consists of a classifier that is trained for each specific source-destination pair using historical shipment, weather, and social media data. The model estimates the transit times for future shipments using Support Vector Machine (SVM). These estimates were validated using four case study routes of varying distances in the United States. A predictive model is trained for each route. The results show that the contribution of each input feature to the predictive ability of the model varies for each route. The mean average error (MAE) values of the model vary for each route due to the availability of testing and training historical shipment data as well as the availability of weather and social media data. In addition, it was found that the inclusion of the historical traffic data provided by INRIXTM improves the accuracy of the model. Sample INRIXTM data was available for one of the routes. One of the main limitations of the proposed approach is the availability of historical shipment data and the quality of social media data. However, if the data is available, the proposed methodology can be applied to any supplier with high volume shipments in order to develop a predictive model for outbound transit time delays over any land route.

In surface hydrology, much attention is paid to the effects of changing water fluxes, however there is less of a focus on the effects of changing energy fluxes. These energy fluxes are an important driver of many hydrological processes such as evapotranspiration and snow sublimation/ablation. The hypothesis that varying energy fluxes are important to the hydrological features of a catchment is tested by an experiment that involves calculating mean transit times for a number of catchments that drain different aspects of a large dome located in the Valles Caldera, New Mexico, called Redondo Peak. These catchments have different orientations and therefore receive different amounts of solar radiation. There is a general correlation between mean transit times, as determined by lumped-parameter convolution, and aspect, suggesting that in the Valles Caldera, transit times might be affected by a variety of features that are influenced by exposure to solar radiation, such as slope steepness, vegetation patterns, and soil depth. To put these transit times into context, I also used a distributed physically-based model to simulate a number of factors simultaneously to determine how hydrological features are influenced by aspect. This modeling excercise has illuminated the aspect-dependence of hydrological features such as the timing and intensity of snowmelt and soil moisture patterns, and it has quantified differences in energy and water fluxes on different aspects. These factors affect both water storage and water fluxes, and are therefore tied to transit times.

On-time delivery of supplies to industry is essential because delays can disrupt production schedules. The aim of the proposed application is to predict transit times for outbound logistics thereby allowing suppliers to plan for timely mitigation of risks during shipment planning. The predictive model consists of a classifier that is trained for each specific source-destination pair using historical shipment, weather, and social media data. The model estimates the transit times for future shipments using Support Vector Machine (SVM). These estimates were validated using four case study routes of varying distances in the United States. A predictive model is trained for each route. The results show that the contribution of each input feature to the predictive ability of the model varies for each route. The mean average error (MAE) values of the model vary for each route due to the availability of testing and training historical shipment data as well as the availability of weather and social media data. In addition, it was found that the inclusion of the historical traffic data provided by INRIX™ improves the accuracy of the model. Sample INRIX™ data was available for one of the routes. One of the main limitations of the proposed approach is the availability of historical shipment data and the quality of social media data. However, if the data is available, the proposed methodology can be applied to any supplier with high volume shipments in order to develop a predictive model for outbound transit time delays over any land route.

Le développement des technologies de communication et de l’information nécessite des composants semi-conducteurs ultrarapides et à faible niveau de bruit. Les transistors bipolaires à hétérojonction (TBH) sont des dispositifs qui visent des applications à hautes fréquences et qui peuvent satisfaire ces conditions. L’objet de cette thèse est l’étude expérimentale et la modélisation du bruit haute fréquence des TBH Si/SiGe:C (technologie STMicroelectronics) et InP/InGaAs (III-V Lab Alcatel-Thales).Accompagné d’un état de l’art des performances dynamiques des différentes technologies de TBH, le chapitre I rappelle brièvement le fonctionnement et la caractérisation des TBH en régime statique et dynamique. La première partie du chapitre II donne la description des deux types de TBH, avec l’analyse des performances dynamiques et statiques en fonction des variations technologiques de ceux-ci (composition de la base du TBH SiGe:C, réduction des dimensions latérales du TBH InGaAs). Avec l’aide d’une modélisation hydrodynamique, la seconde partie montre l’avantage d’une composition en germanium de 15-25% dans la base du TBH SiGe pour atteindre les meilleurs performances dynamiques. Le chapitre III synthétise des analyses statiques et dynamiques réalisées à basse température permettant de déterminer le poids relatif des temps de transit et des temps de charge dans la limitation des performances des TBH. L’analyse expérimentale et la modélisation analytique du bruit haute fréquence des deux types de TBH sont présentées en chapitre IV. La modélisation permet de mettre en évidence l’influence de la défocalisation du courant, de l’auto-échauffement, de la nature de l’hétérojonction base-émetteur sur le bruit haute fréquence. Une estimation des performances en bruit à basse température des deux types de TBH est obtenues avec les modèles électriques. / In order to fulfil the roadmap for the development of telecommunication and information technologies (TIC), low noise level and very fast semiconductor devices are required. Heterojunction bipolar transistor has demonstrated excellent high frequency performances and becomes a candidate to address TIC roadmap. This work deals with experimental analysis and high frequency noise modelling of Si/SiGe:C HBT (STMicroelectronics tech.) and InP/InGaAs HBT (III-V Lab Alcatel-Thales).Chapter I introduces the basic concepts of HBTs operation and the characterization at high-frequency. This chapter summarizes the high frequency performances of many state-of-the-art HBT technologies. The first part of chapter II describes the two HBT sets, with paying attention on the impact of the base composition (SiGe:C) or the lateral reduction of the device (InGaAs) on static and dynamic performances. Based on TCAD modelling, the second part shows that a 15-25% germanium composition profile in the base is able to reach highest dynamic performances. Chapter III summarizes the static and dynamic results at low temperature, giving a separation of the intrinsic transit times and charging times involved into the performance limitation. Chapter IV presents noise measurements and the derivation of high frequency noise analytical models. These models highlight the impact of the current crowding and the self-heating effects, and the influence of the base-emitter heterojunction on the high frequency noise. According to these models the high frequency noise performances are estimated at low temperature for both HBT technologies.