Daimler (United Kingdom)

Visual understanding of complex urban street scenes is an enabling factor for a wide range of applications. Object detection has benefited enormously from large-scale datasets, especially in the context of deep learning. For semantic urban scene understanding, however, no current dataset adequately captures the complexity of real-world urban scenes. To address this, we introduce Cityscapes, a benchmark suite and large-scale dataset to train and test approaches for pixel-level and instance-level semantic labeling. Cityscapes is comprised of a large, diverse set of stereo video sequences recorded in streets from 50 different cities. 5000 of these images have high quality pixel-level annotations, 20 000 additional images have coarse annotations to enable methods that leverage large volumes of weakly-labeled data. Crucially, our effort exceeds previous attempts in terms of dataset size, annotation richness, scene variability, and complexity. Our accompanying empirical study provides an in-depth analysis of the dataset characteristics, as well as a performance evaluation of several state-of-the-art approaches based on our benchmark.

In this paper we formulate structure from motion as a learning problem. We train a convolutional network end-to-end to compute depth and camera motion from successive, unconstrained image pairs. The architecture is composed of multiple stacked encoder-decoder networks, the core part being an iterative network that is able to improve its own predictions. The network estimates not only depth and motion, but additionally surface normals, optical flow between the images and confidence of the matching. A crucial component of the approach is a training loss based on spatial relative differences. Compared to traditional two-frame structure from motion methods, results are more accurate and more robust. In contrast to the popular depth-from-single-image networks, DeMoN learns the concept of matching and, thus, better generalizes to structures not seen during training.

<div class="htmlview paragraph">Chemiluminescence imaging has been applied to a parametric investigation of diesel autoignition. Time-resolved images of the natural light emission were made in an optically accessible DI diesel engine of the heavy-duty size class using an intensified CCD video camera. Measurements were obtained at a base operating condition, corresponding to a motored TDC temperature and density of 992 K and 16.6 kg/m<sup>3</sup>, and for TDC temperatures and densities above and below these values. Data were taken with a 42.5 cetane number blend of the diesel reference fuels for all conditions, and measurements were also made with no. 2 diesel fuel (D2) at the base condition. For each condition, temporal sequences of images were acquired from the time of first detectable chemiluminescence up through fully sooting combustion, and the images were analyzed to obtain quantitative measurements of the average emission intensity. These data are presented and discussed with respect to the liquid-phase fuel penetration and the apparent heat release rates. In addition, a spectrometer equipped with an intensified OMA was used to record both chemiluminescence and sooting combustion spectra.</div> <div class="htmlview paragraph">The data show diesel autoignition to be a progressive process that occurs simultaneously over the downstream region of all the fuel jets. The first chemiluminescence is detected shortly after the start of injection for all the conditions studied, and it grows brighter and shifts downstream as the jet penetrates across the chamber, until soot luminosity dominates the emission. Although this process is qualitatively the same for all conditions, the rate at which it proceeds varies greatly with TDC temperature and density. The temporal evolution of the natural emission is similar for the reference and D2 fuels. The spectral data show that the chemiluminescence arises mainly from formaldehyde and CH emission, with no OH emission detected. In contrast, the sooting-combustion spectrum (taken after the diffusion flame has formed) shows OH and “gray-body” soot emission.</div>

The 2003 International Symposium on Mixed and Augmented Reality was accompanied by a workshop on potential industrial applications. The organizers wisely called it potential because the real use of augmented reality (AR) in an industrial context is still in its infancy. Our own experience in this field clearly supports this viewpoint. We have been actively involved in the research, development, and deployment of AR systems in the automotive, aviation, and astronautics industries for more than five years and have developed and implemented AR systems in a wide variety of environments while working at DaimlerChrysler in Germany. In this article we have selected 10 AR projects from those we have managed and implemented in the past to examine the main challenges we faced and to share some of the lessons we learned.

The cooperative intelligent transport system (C-ITS) (also known as connected vehicle technology in the United States) is an application using vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, at a carrier frequency of 5.9 GHz, to increase road traffic safety and efficiency in Europe. In this article, we hope to shed light on the current status of the C-ITS in Europe and the activities that must be accomplished before deployment can commence in 2019, the date announced by the CAR-2-CAR Communication Consortium (C2C-CC). There is immense activity regarding the launch of the C-ITS in Europe, and the automotive industry is also currently planning for the future.

<div class="htmlview paragraph">A detailed, one-dimensional, time dependend model is presented, describing flame kernel development in spark ignition engines which explicitely accounts for all fundamental properties of the ignition system (supplied electrical energy and power, discharge mode, energy transfer efficiency to spark plasma, plasma temperature distribution, gap width, heat losses to electrodes and chamber walls), of the combustible mixture (pressure, temperature, equivalence ratio, residual gas fraction, laminar burning velocity, type of fuel) and of the flow field (mean flow velocity, turbulence intensity, strain, characteristic time and length scales, flame holder effects). The model is based on the strained flamelet model and predicts kernel growth consistently under virtual all relevant physical/chemical conditions. Model predictions have been verified in extensive studies in an optical engine over a wide range of physical/chemical parameters using advanced optical and laser optical diagnostics. Very good agreement has been obtained between measured and predicted data without a need for adjusting any of the parameters. Four different ranges for flame kernel formation have been identified being governed by the relative magnitudes of plasma velocity, turbulent burning velocity, heat losses, effective strain or flow field effects. The paper describes the derivation of the model, the verification of model predictions against experimental data and the experimental and diagnostic details.</div>

The field of Holonic Manufacturing was initiated in the early 1990's to address the upcoming challenges for manufacturing operations in the 21 century. It is intended to support for highly responsive organisations by providing a modular building-block or "plug and play" capability for developing and operating a manufacturing production system in order to support a more responsive organisation. The holonic approach can be viewed as an alternative to more hierarchical operations management methods such as those based on Computer Integrated Manufacturing (CIM). Since 1990, an increasing amount of research has been conducted in holonic manufacturing over a diverse range of industries and applications, with a strong emphasis on how holonic systems will perform the different planning and control functions required to manage a production operation. The planning and control work to date has, however, been focussed on specific problem formulations and solution strategies. The intention of this paper is to provide an overview of the use of holonic manufacturing concepts in production planning and control which is accessible to both practitioners and researchers in the area. The aims of the paper are: . To clearly define what is meant by a holonic manufacturing system and to demonstrate the relevance of its development to production planning and control.

The Fleetnet project aims at the development of a wireless ad hoc network for inter-vehicle communications. We present the rationale behind the choice of an appropriate radio hardware and the use of a position-based routing approach and outline applications to exploit the Fleetnet platform. In addition, we discuss simulation of vehicle movements as a basis for protocol evaluation as well as aspects of Internet integration of Fleetnet. We state the basic problems together with the intended approach of tackling these challenges, thereby providing an overview of the Fleetnet project

<div class="htmlview paragraph">In this article model-based controller design techniques are investigated for the transient operation of a common-rail diesel engine in order to optimize driveability and to reduce soot emissions. The computer-aided design has benefits in reducing controller calibration time.</div> <div class="htmlview paragraph">This paper presents a nonlinear control concept for the coordinated control of the exhaust gas recirculation (EGR) valve and the variable geometry turbocharger (VGT) in a common-rail diesel engine. The overall controller structure is set up to regulate the total cylinder air-charge with a desired fresh air-charge amount by means of controlling the intake manifold pressure and estimating the fresh air-charge inducted into the cylinders.</div> <div class="htmlview paragraph">During varying engine operating conditions the two control loops are coordinated by a compensation of the EGR valve action through the VGT controller. A nonlinear exhaust pressure controller is designed to regulate the estimated turbocharger power which compensates all EGR valve actions and results in the desired turbocharger power management.</div> <div class="htmlview paragraph">The VGT and EGR controllers, the cylinder air-charge observer and the turbocharger identification algorithm are developed based on a nonlinear diesel engine model. The benefits of the coordinated controller structure are demonstrated with transient engine measurements in a passenger car.</div>

<div class="htmlview paragraph">In modern DI Diesel engines the raw emissions of NOx and soot are affected, apart from the fuel injection rate, by atomization of the liquid jet and mixing of the fuel with the combustion air. Thereby details of the fuel flow inside the injection nozzle play an essential role. In order to determine the general mechanisms and the effect of individual nozzle configuration parameters on the fuel atomization and the fuel spray propagation, methods for optical diagnostics and CFD have been developed at the DaimlerChrysler Research.</div> <div class="htmlview paragraph">These methods are combined with an analysis of the injection system hydraulics and linked to a detailed analysis of mixture formation and combustion inside an optically accessible engine.</div> <div class="htmlview paragraph">The first part of the paper methods for the experimental investigation with transparent 1- and 6- hole nozzles in real size geometries under high pressure conditions are described. Special emphasis is put on CFD methods for modeling the cavitating two phase nozzle flow. In the second part the processes occurring in a sac hole nozzle of a common rail injector during the complete injection event are discussed. The influences of the spray hole position, inlet rounding and conical shape of the spray hole are presented. Finally a comparison between a needle lift controlled (CR = common rail) and a pressure controlled (PLN = pump line nozzle) injection system is made.</div>

<div class="htmlview paragraph">In his paper “The Knock Syndrome - its Cures and its Victims” (SAE <a href="http://www.sae.org/technical/papers/841339" target="_blank">841339</a>) Oppenheim proposed to change the whole process of the internal combustion engine replacing moving flames by homogeneous and simultaneous combustion. Intensive research work on flame propagation and auto-ignition phenomena led to new insights into combustion over recent years. The implementation of auto-ignition on two-stroke S.I. engines revealed the potential for simultaneous reductions in fuel consumption and NOx emission.</div> <div class="htmlview paragraph">Deploying the principle for the four-stroke piston engine and standard fuel would provide optimum conditions for application in common vehicles. The basic problem of homogeneous combustion is presented and some options of control are discussed. A methodology is proposed to apply a new type of combustion simply through a consistent combination of modern technology available for the S.I. engine. The practical investigations and examination of the proposed concept will be realized as part of the joint European research project “4-SPACE”.</div>

Evolutionary structural testing is an approach to automatically generating test cases that achieve high structural code coverage. It typically uses genetic algorithms (GAs) to search for relevant test cases. In recent investigations particle swarm optimization (PSO), an alternative search technique, often outperformed GAs when applied to various problems. This raises the question of how PSO competes with GAs in the context of evolutionary structural testing. In order to contribute to an answer to this question, we performed experiments with 25 small artificial test objects and 13 more complex industrial test objects taken from various development projects. The results show that PSO outperforms GAs for most code elements to be covered in terms of effectiveness and efficiency.

<div class="htmlview paragraph">The effects of in-cylinder water injection on a direct injection (DI) Diesel engine were studied using a computational fluid dynamics (CFD) program based on the Kiva-3v code. The spray model is validated against experimental bomb data with good agreement for vapor penetration as a function of time. It was found that liquid penetration increased approximately 35% with 23% of the fuel volume replaced by water, due mostly to the increase in latent heat of vaporization.</div> <div class="htmlview paragraph">Engine calculations were compared to experimental results and showed very good agreement with pressure, ignition delay and fuel consumption. Trends for emissions were accurately predicted for both 44% and 86% load conditions. Engine simulations showed that the vaporization of liquid water as well as a local increase in specific heat of the gas around the flame resulted in lower Nitrogen Oxide emissions (NOx) and soot formation rates. Using stratified fuel-water injection increases soot at 86% loads due in part to late injection. Because NOx decreased at all loads, the injection timing can be advanced to minimize fuel consumption and soot.</div>

<div class="htmlview paragraph">This paper presents a phenomenological single-zone combustion model which meets the particular requirements of high speed DI diesel engines with common rail injection. Therefore the model takes into account the freely selectable pilot and main injection and is strongly focusing on result parameters like combustion noise or NO-emission which are affected by this split injection. The premixed combustion, the mixing-controlled combustion and the ignition delay are key parts of the model. The model was developed and tested on more than 200 samples from three different engine types of DaimlerChrysler passenger car engines equipped with common rail injection. A user-friendly parameterization and a short computing time was achieved thanks to the simple structure of the model.</div>

The objective of this article is to develop a vehicle sideslip observer that takes the nonlinearities of the system into account, both in the theoretical analysis and the design. The design goals include reduction of the computational complexity compared to the EKF, to make the observer suitable for implementation in the embedded hardware, and a reduction in the number of tuning parameters compared to the EKF. Design is based on a standard sensor configuration, and is subjected to the extensive testing in the realist conditions.

We study the wireless communication among highway vehicles in the newly-assigned 5.9 GHz dedicated short range communication (DSRC) spectrum. A vehicle-vehicle location-based broadcast (LBB) communication protocol is designed to meet highway safety applications' communication requirements. The analytical expressions of the performance of the protocol in terms of probability of transmission failure and channel occupancy are derived with commonly satisfied assumptions. The optimal relation between the performance and design parameters is obtained from the expressions. The sensitivity of the protocol performance is tested for various communication conditions as well as highway traffic conditions. Feasible combinations of the communication and highway traffic parameters are found for certain requirements on protocol performance.

The Fleetnet project aims at the development of a wireless ad hoc network for inter-vehicle communications. We present the rationale behind the choice of an appropriate radio hardware and the use of a position-based routing approach and outline applications to exploit the Fleetnet platform. In addition, we discuss simulation of vehicle movements as a basis for protocol evaluation as well as aspects of Internet integration of Fleetnet. We state the basic problems together with the intended approach of tackling these challenges, thereby providing an overview of the Fleetnet project

<div class="htmlview paragraph">Predictive Cruise Control (PCC) is a system that enhances and works in combination with the existing Conventional Cruise Control. Based on elevation information captured in a 3D map and a predictive algorithm, PCC allows the vehicle speed to vary around the cruise control set speed within a defined speed band in an effort to reduce fuel consumption. As fuel consumption is a major portion of a truck's life cycle costs (LCC) and cruise control is used extensively in the United States and Canada, PCC can significantly reduce the truck's LCC.</div>

We seek to design an AR interface that provides users with interactivity so rich it would merge the physical space in which we live and work with the virtual space in which we store and interact with digital information. In this single augmented space, computer-generated entities would become first-class citizens of the physical environment. We would use these entities just as we use physical objects, selecting and manipulating them with our hands instead of with a special-purpose device such as a mouse or joystick. Interaction would then be intuitive and seamless because we would use the same tools to work with digital and real objects. Tiles is an AR interface that moves one step closer to this vision. It allows effective spatial composition, layout, and arrangement of digital objects in the physical environment. The system facilitates seamless two-handed, three-dimensional interaction with both virtual and physical objects, without requiring any special-purpose input devices.

<div class="htmlview paragraph">The overall program objectives were three fold:</div> <div class="htmlview paragraph"> <ul class="list disc"> <li class="list-item"><div class="htmlview paragraph">assess the benefits and limitations of oxygenated diesel fuels on engine performance and emissions</div></li> <li class="list-item"><div class="htmlview paragraph">identify oxygenates most suitable for potential use in future diesel formulations based on physico-chemical properties (e.g. flash point), toxicity, biodegradability and estimated cost of production</div></li> <li class="list-item"><div class="htmlview paragraph">perform limited emissions and performance testing of the oxygenated diesel blends</div></li> <li class="list-item"><div class="htmlview paragraph">select at least two oxygenated compounds for advanced engine testing</div></li> </ul> </div> <div class="htmlview paragraph">In Part 1 of this program which is described in this paper, an extensive literature review was conducted to identify potential oxygenates for blending into diesel fuels. As many as 71 oxygenates were identified for the initial screening process. Based on a set of physical and chemical properties, a screening methodology was developed to select the 8 oxygenates that will be eligible for engine testing. The following properties were used:</div> <div class="htmlview paragraph"> <ul class="list disc"> <li class="list-item"><div class="htmlview paragraph">oxygen content, flash point, solubility in low aromatic diesel fuel and stability</div></li> <li class="list-item"><div class="htmlview paragraph">fuel properties such as viscosity, cetane number and lubricity</div></li> <li class="list-item"><div class="htmlview paragraph">corrosivity</div></li> <li class="list-item"><div class="htmlview paragraph">elastomer compatibility</div></li> <li class="list-item"><div class="htmlview paragraph">toxicity</div></li> <li class="list-item"><div class="htmlview paragraph">biodegradability</div></li> </ul> </div> <div class="htmlview paragraph">Initial screening resulted in 43 candidate oxygenates. Properties not found in the literature such as solubility and corrosivity were evaluated using bench tests. These tests resulted in the selection of eight oxygenates for engine testing in Part 2 of this program. These oxygenates were blended with alternative low sulfur diesel (ALS) to achieve an oxygen content of 7 wt%. Tables showing oxygenate properties are included.</div>