In this paper, we examined the influence of relatively advanced voice control task which are many conversation hierarchies and can respond to narrowing-down searches tailored to driver's preferences and utterances of different phrases with the same meanings. We have set the factor for cognitive complexity that are common performance to any voice control task, and have investigate its influence on the driving. As the result, voice-control interaction is clearly smaller variability lane position and better subjective performance and better response ability of sudden events on forward road, while is same headway variability relative to the same task using a visual-manual interaction. This result is useful for designing automotive voice-control system.

View full abstract

The goal of this study is to clarify a method to provide safety-driving information which can induce behavioral changes by increase of intrinsic motivation with verbal rewards. First, component of cognition and emotion which is elicited by the verbal positive feedback was investigated through an internet survey. The result indicated that the motivation by the verbal positive feedback consisted of 4 factors. Next, it was examined whether increasing the motivation by the positive feedback can improve visual behavior in simulated driving scene. Many of the participants showed improvement of eye movement to hazard by the verbal rewards including informative element.

View full abstract

In this paper, balancing motion of drivers are analyzed using characteristics of angular momentum about center of gravity. In the analysis, driver’s body is instantaneously assumed as upper and lower body. Then the angular momentum vectors of the two bodies during driving maneuver respectively decomposed into the steering wheel rotational component and the orthogonal component in the sagittal plane. A measure of mutual cancellation is employed to analyze motion of six participants in different skill levels at slalom maneuver. Also characteristics while walking and running are analogously compared. The results give insights to explain drivers’ feeling on hold performance, and show potential of angular momentum as a control variable in driver’s motion in-vehicle.

View full abstract

The aim of this study is to understand the mechanism of dew condensation and try to prevent it on automobile headlamp. In this case, it is necessary to accurately reproduce the complicated shape of the headlamp as the computational model. In addition, the simulation framework for predicting the turbulent flow field accompanied by the high heat source and the wall heat transfer with high accuracy in the vehicle running environment is required. In this paper, we report that we tried constructing the simulation method of compressible turbulence based on hierarchical Cartesian grid using the HPC (High Performance Computing) environment.

View full abstract

Numerical analysis has been performed for vertical buoyant turbulent flow of air-helium mixture plume in open space by using algebraic Reynolds stress model and algebraic turbulent mass flux model. Calculated results are compared with the experimental data of velocity profile, helium concentration and Reynolds stresses in order to confirm the validity of the presented algebraic turbulent mass flux model. Although the agreement is certainly not perfect especially for vertical normal stress, the main features, which are accelerated flow, rapid diffusion of helium and periodic puff cycles induced by Rayleigh-Taylor instability, are reasonably predicted by the presented algebraic turbulent mass flux model.

View full abstract

In recent years, car users have been provided with a service for managing fuel economy data by posting the fuel supply and travel distance to the WEB at every refueling. If there is no mistake in the contribution data, actual fuel economy will be given to the user. For an analyst trying to utilize a lot of refueling data, some consideration is required for the reliability of the data and the statistical bias in the contributors. In this research, we report the results of actual fuel economy analysis using user contributed data for one year of micro vans nationwide.

View full abstract

In order to prevent air pollution and to reduce CO₂ emissions, Electric Vehicles (EVs) that do not emit exhaust gas while driving are beginning to spread worldwide. But considering exhaust emissions from power plants, exhaust gas dose not be said to be zero. In this paper, the author focus on CO₂ emissions from power plants, evaluate the CO₂ emissions from EVs and discuss the surrounding conditions of EVs around 2030 for reducing more CO₂ emissions than hybrid vehicles.

View full abstract

Some ultra-high-strength steel frames with hat-shaped section and jointed flange by spot welding were given 4 point bending deformation. The validity of 4 point bending test system was confirmed by some simulation. The test system could give the frames constant moment in the wide range that was characteristic of 4 points bending condition. Also, by static and dynamic 4 point bending test, in case of the frames with short flange in cross direction, a crack occured in heat affected zone around spot welding point on the flange of the frames had fracture. On the other hand, the frames with large flange did not have fracture in the heat affected zone.

View full abstract

Measured residual stress is affected by the friction between steel sheet and electrode during welding. However, effects of friction on residual stress are not considered in numerical simulation. Therefore, it is expected that the experiment results don’t coincide with numerical simulation results. In this study, welding condition was proposed to produce the joints without effect of friction on residual stress. Moreover, the effects of welding condition on residual stress distribution were investigated. These results indicated that increasing temperature and decreasing pressure on sheet surface were necessary to produce the joints without effect of friction.

View full abstract

The high metallic silver using thin smooth aluminum flake pigment is placed as the luxury grade in each car manufacturer. But it has a technical difficulty, the color mottling by unevenness of the pigment distribution and several problems is also easy to occur. Therefore the design that we could reach had a limit. In this article we would like to explain the new type low non volatile water borne paint system using cellulose nano fiber, which can provide the high metallic color stably without color unevenness, even when the ultrathin aluminum pigment such as the vapor deposition aluminum.

View full abstract

There are many production robots used at car manufacturing plants, and each of them uses several reducers. A breakdown of one of these reducers may cause a huge loss due to the stop of all production lines. Therefore, the condition based maintenance to predict failures by predetermined thresholds of average and standard deviation is being currently used. In this study, we propose a new failure prediction method using probability density function and noise removals to improve the accuracy of prediction. Evaluations of prototypes showed much better results and its performance was applicable to the actual production lines.

View full abstract

Although hammering sheet-metal processing is a typical incremental metal forming process in auto-repair shop, it has not still been mechanized because of its high skilled techniques with its intermittent high nonlinear properties. In this study, we applied a dead beat transfer function called vibration manipulation function (VMF) to command the hammering forming process under piecewise sampled-data control. From numerical calculations, we found this method can control a hammering machine with linear spring under various conditions. Although it can handle a pendulum-type hammering machine with small excitation, the method will be failed with large amplitude, except in certain operational period under small coefficient of restitution.

View full abstract

To control next generation engine, it is required to predict state of in-cylinder gas. We have obtained gas flow rate around intake valve to improve the prediction. High-response laminar flow meter was used to measure transient gas flow. And, the flow rate around intake valve was calculated by the model of laminar flow meter and intake pipe. This paper shows the evaluation of obtained gas flow rate and integrated charging efficiency.

View full abstract

The complexity of electronic control systems in automobiles are increasing. AUTOSAR software standard was introduced to develop increasing complexity by layered architecture and standardized components. In safety-critical systems, the diagnostic software consists of a mixture of a cyclic execution and a hardware-event-driven. Testing the multiple concurrent processes is impractically time consuming. We propose the layered architecture of verification system, which complies to the AUTOSAR abstraction layers and how to model a combination of the concurrent executions of multiple tasks under non-deterministic event from hardware on CBMC as a verification engine and made an extension to CSeq as a concurrency pre-processor.

View full abstract

This paper presents a reliability estimation method of localization results. In the method, an egovehicle pose and reliability are treated as hidden variables and are estimated simultaneously via Rao- Blackwellized particle filter (RBPF). The ego-vehicle pose is estimated by a sampling-based method, i.e., particle filter, and the reliability is estimated by an analytical method using prediction results of convolutional neural network (CNN). The CNN learns whether localization has failed or not and its output is used as an observable variable to estimate the reliability in the RBPF. Through experiments, it is shown that the estimated reliability could be used as an exact criterion for describing successful and fault localization results.

View full abstract

Heavy-duty hybrid electric vehicles (HEV) and Heavy-duty plug-in hybrid electric vehicles (PHEV) are anticipated an increase in number. Fuel economy and emission for Heavy-duty HEV is evaluated by using HILS (Hardware-In-the-Loop-Simulator). Engine part of HILS is built by engine torque map and fuel consumption map. These maps are obtained at the engine hot condition. In the future, a new evaluation method of fuel economy and emission for HEV and PHEV is necessary. The reason is that these vehicles are driven at engine cold start condition. In this study, virtual HEV was estimated by using Engine-In-the-Loop, which is combined actual engine and HILS. Powertrain performance, fuel economy and emissions for virtual HEV was evaluated, and compared with actual HEV by using chassis dynamometer.

View full abstract

In this paper, we build a deep reinforcement learning that outputs the optimum speed acceleration to improve fuel economy while taking account into traffic flow in public highway. The deep reinforcement learning utilizes the distance of spatiotemporal feature of traffic signal condition. From the simulation evaluation of our scheme, we could confirm that fuel efficiency can be improved by about 10% while maintaining the overall traffic flow in the proposed method, against the rule-based speed adjustment scheme.

View full abstract

The authors are operating a database system for collecting traffic near-miss incident cases. We propose a new annotation method suitable for computer data analysis while inheriting existing data sets. We also design software interfaces for realizing secure and efficient data access. In this study, the evaluation experiments of the efficiency of the proposed database system are reported.

View full abstract