No abstract

No abstract

No abstract

No abstract

Abstract-We address the vehicle detection and classification problems using Deep Neural Networks (DNNs) approaches. Here we answer to questions that are specific to our application including how to utilize DNN for vehicle detection, what features are useful for vehicle classification, and how to extend a model trained on a limited size dataset, to the cases of extreme lighting condition. Answering these questions we propose our approach that outperforms state-of-the-art methods, and achieves promising results on image with extreme lighting conditions.

This thesis examines the vehicle design and sales mix changes necessary to double the average fuel economy of new U.S. cars and light-trucks by model year 2035. To achieve this factor of two target, three technology options that are available and can be implemented on a large scale are evaluated: (1) channeling future vehicle technical efficiency improvements to reducing fuel consumption rather than improving vehicle performance, (2) increasing the market share of diesel, turbocharged gasoline and hybrid electric gasoline propulsion systems, and (3) reducing vehicle weight and size.The illustrative scenarios demonstrate the challenges of this factor-of-two improvement --major changes in all these three options would need to be implemented before the target is met. Over the next three decades, consumers will have to accept little further improvements in acceleration performance, a large fraction of new light-duty vehicles sold must be propelled by alternative powertrains, and vehicle weight must be reduced by 20-35% from today. The additional cost of achieving this factor-of-two target would be about 20% more than a baseline scenario where fuel consumption does not change from today's values, although these additional costs would be recouped within 4 to 5 years from the resulting fuel savings.Thus, while it is technically feasible to halve the fuel consumption of new vehicles in 2035, aggressive changes are needed and additional costs will be incurred. Results from this study imply that continuing the current trend of ever increasing performance and size will have to be reversed if significantly lower vehicle fuel consumption is to be achieved.

No abstract

SummaryIn this article, a methodology to model the annual stock and flows of aluminum in a key end-use sector in the United States-passenger vehicles-from 1975-2035 is described. This dynamic material flow model has enabled analysis of the corresponding energy embodied in automotive aluminum as well as the cumulative aluminum production energy demand. The former was found to be significant at 2.6 × 10 9 gigajoules (GJ) in year 2008 under baseline assumptions. From 2008-2035, the cumulative energy required to produce aluminum to be used in vehicles is estimated at 7.8 × 10 9 GJ. Although the automotive aluminum stock is expected to increase by 1.8 times by 2035, the corresponding energy embodied is not expected to grow as rapidly due to efficiency improvements in aluminum processing over time. The model's robustness was tested by checking the sensitivity of the results to variations in key input assumptions, including future vehicle sales, lifetimes, and scrap recovery. Sensitivity of energy embodied in automotive aluminum to changes in aluminum production efficiency and aluminum applications within the vehicle were also explored. Using more recycled aluminum or improving the energy efficiency of aluminum production at a faster rate can lower production energy demands. However, aggressive and sustained changes are needed beginning today to achieve meaningful reductions. This may potentially be countered by increased use of stamped aluminum in vehicles.