Search Results

More than four decades ago, the concept of zero defects was coined by Phillip Crosby. It was only a vision at the time, but the introduction of Artificial Intelligence (AI) in manufacturing has since enabled it to become attainable. Since most mature manufacturing organizations have merged traditional quality philosophies and techniques, their processes generate only a few Defects Per Million of Opportunities (DPMO). Detecting these rare quality events is one of the modern intellectual challenges posed to this industry. Process Monitoring for Quality (PMQ) is an AI and big data-driven quality philosophy aimed at defect detection and empirical knowledge discovery. Detection is formulated as a binary classification problem, where the right Machine Learning (ML), optimization, and statistics techniques are applied to develop an effective predictive system.

Understanding customer expectations is critical to satisfying customers. Holding customer clinics is one approach to set winning targets for the engineering functional measures to drive customer satisfaction. In these clinics, customers are asked to operate and interact with vehicle systems or subsystems such as doors, lift gates, shifters, and seat adjusters, and then rate their experience. From this customer evaluation data, engineers can create customer loss or preference functions. These functions let engineers set appropriate targets by balancing risks and benefits. Statistical methods such as cumulative customer loss function are regularly applied for such analyses. In this paper, a new approach based on the Taguchi method is proposed and developed. It is referred to as Taguchi Customer Loss Function (TCLF).

Due to the limitations on resistance spot welding of ultra-thin steel sheet (thicknesses below 0.5 mm) in high-volume automotive manufacturing, a comparison of friction stir spot welding and self-piercing riveting was performed to determine which process may be more amenable to enabling assembly of ultra-thin steel sheet. Statistical comparisons between mechanical properties of lap-shear tensile and T-peel were made in sheet thickness below 0.5 mm and for dissimilar thickness combinations. An evaluation of energy to fracture, fracture mechanisms, and joint consistency is presented.

Understanding a system behavior involves developing an accurate relationship between the explanatory (predictive) variables and the output response. When the observed data is ill-conditioned with potential collinear correlations among the measured variables, some of the statistical methods such as least squared method (LSM) fail to generate good predictive models. In those situations, other methods like Principal Component Regression (PCR) are generally applicable. Additionally, the PCR reduces the dimensionality of the system by making use of covariance relationship among the variables. In this paper, an improved regression method over PCR is proposed, which is based on the Trivial Principal Components (TPC). The TPC regression (TPCR) makes use of the covariance of the output response and predictive variables while extracting principal components. A new method of selecting potential principal components for variable reduction in TPCR is also proposed and validated.

The use of design of experiment (DoE) and data-driven simulation has become state-of-the-art in engine development and base calibration to cope with the drastically increased complexity of today's engine ECUs (electronic control units). Based on the representation of the engine behavior with a virtual plant model, offline optimizers can be used to find the optimal calibration settings for the engine controller, e.g. with respect to fuel consumption and exhaust gas emissions. This increases the efficiency of the calibration process and reduces the need for expensive test stand runs. The present paper describes the application of Gaussian process regression, a statistical modeling approach with practical benefits in terms of achievable model accuracy and usability. The implementation of the algorithm in a commercial tool framework enables a broad use in series engine calibration.

Robust engineering is an integral part of the quality initiative, Design For Six Sigma (DFSS), in most companies to enable good designs and products for reliability and durability. Taguchi’s signal-to-noise ratio has been considered as a good performance index for robustness for many years. An alternate approach that is direct and simple for measuring robustness is proposed. In this approach, robustness is measured in terms of an augmented output response and it is a composite index of variation and efficiency of a system. This formulation represents an engineering design intent of a product in a statistical sense, so engineers can understand, communicate, and resonate at ease. Robust formulations are illustrated and discussed with case studies for smaller-the-better, nominal-the-best, and dynamic responses. Confirmation runs of optimization show good agreement of the augmented response with the additive predictive models.

Electric motor whine is one of the main noise sources of hybrid and electric vehicles. Motor air gap eccentricity due to propulsion system deflection, part tolerances and manufacturing variation is typically ignored in motor NVH design and analysis. Such eccentricity can be a dominant noise source by amplifying critical motor whine orders up to 10 dB, leading to poor NVH robustness. However, this problem cannot be explained by conventional method based on symmetric 2D approach. New 3D electromagnetic (EM) and NVH analyses are developed and validated to accurately predict air gap induced motor noise to enhance NVH robustness: First, a true 3D full 360-degree electric motor model is developed to model asymmetric air gap distribution along motor stack length. Predicted 3D EM forces are mapped to mechanical finite-element mesh over the cylindrical stator surface.

The REAL NOx regulation requires tracking and reporting of NOx emissions starting in 2022MY for both medium-duty and heavy-duty diesel vehicles with potential to be considered during the next light-duty rulemaking. The regulation includes minimum NOx mass measurement accuracy requirements of either +/−20 percent or +/− 0.1 g/bhp-hr. Existing NOx sensor technology may not be able to meet the regulated accuracy requirements especially when exposed to other sources of variation within the emissions control system. This paper provides an assessment of real-world NOx sensor accuracy and the impact of other sources of variation and noise factors on NOx measurement accuracy. Noise factors investigated include NOx sensor tolerance, exhaust flow rate estimation, NOx sensor ammonia (NH3) cross sensitivity, mass air flow (MAF) sensor accuracy, NOx sensor placement, and laboratory emissions measurement capability.

The proposed Euro 7 regulation includes On Board Monitoring, or OBM, to continuously monitor vehicles for emission exceedances. OBM relies on feedback from existing or additional sensors to identify high emitting vehicles, which poses many challenges. Currently, sensors are not commercially available for all emissions constituents, and the accuracy of available sensors is not capable enough for in use compliance determination. On board emissions models do not offer enough fidelity to determine in use compliance and require new complex model innovation development which will be extremely complicated to implement on board the vehicle. The stack up of multi-component deterioration leading to an emissions exceedance is infeasible to detect using available sensors and models.

The subsystem of front of dash (FOD) and instrument panel (IP) is a critical path to isolate the powertrain noise and road noise for vehicles. This subsystem mainly consists of sheet metal, dash mats, IP, and the components inside IP such as HVAC and wiring harness. To achieve certain level of cabin quietness, the sound transmission loss performance of this subsystem is usually used as a quantifier. In this paper, the sound transmission loss through the FOD and IP is investigated up to 10kHz, through both acoustic testing and numerical simulation. In the acoustic testing, the subsystem is cut from a vehicle and installed on the wall of two-rooms STL testing suite, with source room being reverberant and receiver room being anechoic. In the testing, various scenarios are measured to understand the contributions from different components.