Global Automotive + Manufacturing Engineering MEng Projects

The application of additive manufacturing (AM), also known as 3D-printing (3DP), has been introduced not only to the engineering areas, but also to medical and clinical treatments. The material extrusion (MEX) process, as one of the most commonly used AM technologies, is widely adopted to fabricate customized 3D-printed medical devices. This research discusses the feasibility of implementing an innovative MEX passive dynamic ankle-foot orthosis (AFO), denoted as the 3DP AFO, in China. China's orthopedic orthotics devices industry size was approximately USD 6.7 billion (RMB 45 billion), in 2018. Researches and data show that diabetes, obesity, stroke, rheumatoid arthritis and accidents are the leading causes of walking disability. The prevalence of these diseases has become a worldwide health. By applying 3D-printing, low-cost, high material deposition rate, and no powder and laser are key values of using MEX in clinics, 3DP can also bring up to optimization and simplification of the production process, as well as customers’ overall satisfaction in clinical service. With the increasing demand for orthopedic devices and continuous development of the national economy, the penetration of orthopedic devices in China is expected to have a quick lift. An ideal profit model after implementing 3DP AFO is also expected in China.

Altair Engineering Inc. through its technology that optimizes the analysis, management and visualization of business and engineering information wants to try and solve the ‘Design Paradox’ to make the design process of developing a product more robust and efficient. Altair wants to solve this paradox by developing simulation-based technologies that will allow you to make informed decisions and optimize the designs of the product in its early stages. The goal of this project is to optimize the Design Process of an electric vehicle by studying the fan shroud study of the Altair Arc electric vehicle. This kind of Design Process allows you to study the mass flow through the fan shroud and evaluate the drag and lift values for the vehicle even during the initial concept generation stage of the vehicle.

This design process story of the fan shroud shows the implications of using simulation driven techniques and methodology in any design process of a product. It is seen that a simulation driven design in early stages of the process allows you to eliminate any losses that occur due to design problems and helps in coordinating any compromises that occur due to it, which in turn helps in making the whole process more robust and efficient.

In the consumer space, bulk data collection and analysis has been around for quite some time. In the report entitled “Automotive Quality Development Using Connected Customer Bulk Data Analysis” our team’s goal was to explore how the analysis of bulk customer vehicle data can lower warranty and increase customer satisfaction. On average, automotive OEMs spend $9 billion per year on warranty so even small improvements in this area can lead to substantial savings for the manufacturers. Through the use of two major case studies, or team evaluated a process to mitigate warranty in two major categories; customer complaints and warranty spills which make up roughly 35% of total warranty, or $3.15 billion. During the duration of this project, our team has identified areas in which this process can be improved for future users, as well as a business unit within General Motors that can own and refine it. Once rolled out to the proper internal engineering users it is estimated that this process can conservatively mitigate one percent, or $100 million, of customer warranty.

In the consumer space, bulk data collection and analysis has been around for quite some time. In the report entitled “Automotive Quality Development Using Connected Customer Bulk Data Analysis” our team’s goal was to explore how the analysis of bulk customer vehicle data can lower warranty and increase customer satisfaction. On average, automotive OEMs spend $9 billion per year on warranty so even small improvements in this area can lead to substantial savings for the manufacturers. Through the use of two major case studies, or team evaluated a process to mitigate warranty in two major categories; customer complaints and warranty spills which make up roughly 35% of total warranty, or $3.15 billion. During the duration of this project, our team has identified areas in which this process can be improved for future users, as well as a business unit within General Motors that can own and refine it. Once rolled out to the proper internal engineering users it is estimated that this process can conservatively mitigate one percent, or $100 million, of customer warranty.

Fuel cells are a critical next-generation energy technology that will reduce costs and boost efficiencies across several industries. The manufacturing processes of fuel cells involve stacking numerous thin plates of different material to construct a fuel cell stack. One of the assembly strategies that can be used during this assembly process is to locate the plates and cells on pins, or rods. However, a consistent issue with this approach is that plates can bind on the rods and will not assemble. This condition is known as “rod-lock”.

This project focused on studying the interaction between the fuel cell plates and the rods. Through experimentation and data analysis it was recommended that the geometrical tolerances for the plate and rods should be modified in such a way to minimize friction between the two. The expected value-add is in the improved quality and throughput of the manufacturing process, which directly reduces the cost of production.

After a significant retooling of general assembly for a new model vehicle, a throughput analysis of general assembly was performed to identify top bottlenecks. Using throughput metrics including Mean Time to Repair, Mean Cycles Before Failure and Stand Alone Availability, two of twelve lines were identified as significant bottlenecks. After simulating a rebalance of both lines as well as remedying top alarms with software changes, a predicted throughput savings of over $130,000 annually was achieved. The bottlenecks had their throughput improved by approximately 20-30% each. Additionally, work was done to validate that actual conveyor speed matched the desired speed. Lastly, new software was configured to improve access and readability of downtime information for leaders on the plant floor. This allows better response to top issues with readily available and accurate information.

Investigation into warranty claim and consumer metric survey data for General Motors (GM) vehicles from 2016-2019 revealed an opportunity to improve customer satisfaction by enhancing current detection modes for seat cushion wrinkles during the development cycle. Existing GM seat durability trim cover test requirements focus primarily on catastrophic damage at the end of life. A comprehensive customer survey completed for this project identified a high correlation between customer unacceptability of wrinkles and the potential for warranty claims. This suggests a need for a requirement representative of short time in service users. In fact, the survey identified clear thresholds of wrinkle unacceptability that can be mapped in strategic increments throughout the warranty period. With this evidence, the team has proposed two new wrinkle requirements to be evaluated during the seat trim durability test – one at the representative cycle of a new customer, and one at the representative cycle of the end of warranty. We estimate that 80% of the warranty costs associated with seat bolster wrinkles could be avoided if design controls are adopted to meet the proposed requirements. The other 20% is attributed to warranty claims resulting from use cases or defects outside the scope of GM requirements.

In addition to examining the requirements for seat trim durability, we wanted to optimize the life cycle requirement of the test itself. Using GM’s OnStar database, we found that customers are completing 22% less cycles than the current ingress/egress test requires over the course of a GM vehicle’s life. This suggests that the seat ingress/egress durability life cycle requirement may be reduced 22% without risk of missing current detection modes, which translates to a savings of $10,400 in test costs and 2 weeks in test time for a driver seat alone during a typical validation development cycle. Furthermore, this savings may be extended to the rest of the seats in the vehicle.

The project provides experimental design and interface requirement recommendations to improve the sound transmission loss and leakproof performance of the automotive Heating Ventilation and Air Conditioning interface to the vehicle front of dash, through the Design for Six Sigma Identify, Define, Develop, Optimize, and Verify methodology.

Three experimental designs are recommended to support dimensional optimization, seal material and manufacturing process optimization, and verification of optimal results. By determining and utilizing a standard optimized design, General Motors is avoiding potential retrofit costs of $240,000 per year.

Environmental regulation and consumer and manufacturer eco-consciousness necessitates improvement in fuel economy, large vehicles must make use of fuel-saving technology to maintain sales and profitability. One increasingly ubiquitous technology is Automatic Vehicle Start Stop (AVSS), which shuts down the engine when the vehicle is not moving (such as at a stoplight). The goal for manufacturers such as GM is to make the system as “invisible” as possible, primarily by reducing noise and vibration. The goal of this project will be to make use of an Advanced Development Vehicle (ADV) prototype of a 12v mild-hybrid system and perform development engine calibration work to improve it to near-production level. This means that the system will be equal to or better than current production 12v AVSS for vehicle performance and will improve significantly on noise reduction.

After benchmarking current GM and competitor full-size vehicles, and performing calibration work on the prototype vehicle, it is recommended that development of the SGU prototype system continue for the next generation of full-size trucks. Not all the original goals of the project could be met, such as the complete benchmarking of the prototype and competitive testing of the Ram eTorque mild-hybrid system. But the results of testing and analysis that were completed show a need for noise reduction in GM systems versus competitor vehicle (need for 10% noise reduction in 3000-8000Hz range), and significant noise reduction in the prototype system versus the current production GM system (30-50% noise reduction in 1000-6000Hz range). Because of inability to complete all goals, several have been moved into recommendations for continuous work on this project. The ADV team will continue to optimize this system.

The Virtual Development, XIL, and Chassis Controls team within General Motors has a simulation model that uses a combination of Model, Software, Hardware, and Driver-In-the-Loop (i.e., XIL) technologies for development and validation of vehicle dynamics control systems, including Anti-Lock Brakes (ABS) and Electronic Stability Control (ESC). This project explored the feasibility of modifying the existing simulation framework to support model-based systems engineering synthesis at early stages of vehicle development. Two methods of model change were investigated and one proved successful for modification of existing tools to build a synthesis model. The synthesis model is designed to serve multiple objectives within the virtual development, systems engineering, and advanced vehicle development teams. This modified tool was correlated to physical vehicle data and HIL simulation to ensure data integrity, and finally used in a synthesis in support of selection and optimization of vehicle dynamics subsystems for a proposed future vehicle. This synthesis model directly supports the elimination of two prototype vehicles that would have to be built in absence of this simulation capability. Indirectly, three subsystem hardware test benches, or potentially three more prototype vehicles for early learnings, may have been eliminated by the data generated in this report.

General Motors (GM), a major domestic automotive manufacturer, utilizes multiple suppliers for exhaust flex couplers and each supplier uses their own Noise and Vibration (N&V) modeling methods during their in-house product development which hinders GM’s N&V performance characteristic comparison between suppliers. The goal of the project is to investigate various modeling methods for exhaust flex couplers and compare computer-aided engineering N&V accuracy against real-world results. Having an accurate N&V modeling method, we expect to be able to reduce total required tests and eliminate late program timing changes. For each issue with the exhaust flex coupler, we anticipate saving 3-5 physical tests which totals up to $50,000 in testing costs.

This project focused on a front-wheel drive crossover utility vehicle application with the intention of applying the learnings of the project to other vehicle programs. We developed and ran N&V models consistent with industry best practices and performed component testing and in-vehicle physical testing for comparison. Frequency response analysis shows that the three model types that were tested respond similarly to each other but differ from the in-vehicle test results. All three flex coupler models show far more isolation from one end of the flex coupler to the other when compared to in-vehicle testing. Due to these results, continuation of additional testing and model modifications are recommended before implementing any of the new modeling methods.

The Transportation as a Service industry is growing rapidly, with potential for General Motors to capitalize on through the introduction and implementation of Autonomous Vehicles. One of the many hurdles in using Autonomous Vehicles for ride sharing is maintaining a vehicle that is perceived to be clean by customers. Previously conducted research shows that 58% of users will avoid future use of a rideshare service due to poor perceived quality, representing a potential revenue loss $20 million to $2.1 billion.

Odor is a leading indicator of cleanliness in a vehicle. Fecal matter, vomit, body odor and smoke have been found to be most displeasing, and actionable, from a customer’s perspective. This project will focus on the smoke element; specifically, 2^nd and 3^rd hand smoke from cigarettes and e-cigarettes. The main objectives for this project are to identify existing off-the-shelf sensor solutions, and evaluate if it is feasible to detect 2^nd and 3^rd hand smoke conditions with these sensors.

Currently some of the steps in the release process of Software and Calibration datafiles for GM Propulsion Systems (Transmission and Engine) is manually performed by Propulsion Control System Integration Engineers (PCSIEs). They are expending a significant amount of time gathering, comparing and analyzing information from some GM systems and databases to ensure the correctness of the release of part numbers in time to 27 vehicle assembly plants around the world. The purpose of this project is to develop a software tool that assists the PCSIE in doing some of these manual activities automatically, thus reducing the execution time and potential for human errors in these process steps. The project was completed successfully in time and pilots were run with the new PCSIE software tool that resulted in an efficiency improvement of about 29.6%, which translates to about 5,877 engineer hours saved annually corresponding to 18 PCSIEs spread in both GM Mexico and GM USA.

Company ElringKlinger Motortechnik, an engineering supplier in Germany, is facing an increasing demand of testing services focusing on vehicle emissions. Goal of this project was to evaluate an appropriate measurement equipment for emission testing according to RDE-legislation in order to support customers in the automotive industry during the development process and hence, increase revenue of the company. Two systems have been chosen, a portable emission measurement system (PEMS) from AVL as well as a high voltage device from Klaric. The decision was supported by a business case calculation, showing a return on investment within 12 months. A test campaign with a conventional diesel-vehicle and a plug in hybrid was performed in order to evaluate the quality of the measurement equipment as well as to understand the performance of hybrid powertrains. The tests have been done on public roads according to RDE-requirements as well as on a roller dyno to determine the CO₂-emissions according to WLTP. The results were showing an advantage towards the plug-in hybrid, especially in urban areas. However, when considering the emissions caused by electrical grid, the hybrid is close to the emission level of the diesel-vehicle. A strong discrepancy has been observed between RDE and WLTP-results due to the artificial calculation method of the WLTP-procedure which leads to significantly lower values. A small simulation study was done in addition in order to assess the potential of technical modifications to meet the future emission targets in 2030. The new measurement equipment will be a cornerstone for future test campaigns at ElringKlinger. Nonetheless, further improvements are recommended in order to increase quality of the results as well as repeatability.

The future of transportation is filled with electric and autonomous vehicles that will make roads safer, give people back the time they now spend driving, and reduce congestion and emissions. Deploying all-electric autonomous vehicles either in a rideshare environment or for personal use will be heavily dependent on electric vehicle (EV) infrastructure. This project is intended to provide a recommendation of countries, and cities in those countries, that are likely to be large adopters of EV technology. To select countries of interest various factors like population, EV adoption rate, existing EV charging infrastructure, current and future EV legislation, charging standards, and the total number of chargers available were analyzed. Six countries of interest were selected based on this research and initial conclusions and they are China, United States, Germany, Netherlands, Japan and Norway. Once these countries were selected, cities in each of these countries could be selected as cities of interest. The selection of these cities was based on population and they are Shanghai, Beijing, New York City, Los Angeles, Berlin, Hamburg, Amsterdam, Rotterdam, Tokyo, Yokohama, Oslo and Bergen. These recommendations will allow Cruise Automation to expand quickly into the cities with the highest potential. This will give Cruise Automation a business advantage, resulting in international autonomous vehicle deployment and ultimately making roads around the world safer.

The purpose of this project is to analyze the machine downtime, material shortages from global supply chain or machining, and operator overcycles in order to produce xx.x jobs per hour.

The techniques used were data driven from Production Monitoring and Controls to analyze each cycle throughout the day and utilized the DMAIC process through the focus tree to root cause issues and drive them to be fixed. The focus tree was crucial to awareness on the floor as well as changing the operator’s mindset to bringing up their issues so that they get resolved.

The main purpose of this project is the creation of a first-order model for automotive HVAC ducts. The model aims to deliver reasonably accurate assessments of the performance of any given automotive HVAC system based on its geometry and individual characteristics, providing engineers a systematic approach to begin the process of duct sizing while also understanding how component-level decisions affect overall vehicle-level functional requirements.

The opportunity identified is that during the early stages of any vehicle program, HVAC engineers do not have access to detailed CFD (Computational Fluid Dynamics) analysis or physical test data to support their position during the delicate vehicle packaging negotiations. The models described herein provide a simple tool for engineers to use with a sufficient degree of confidence during these discussions. The project success will be measured by the degree to which the tool is adopted by the HVAC ducts Global Subsystem Leadership Team (GSSLT) as standard work to size ducts and assess the impact of packaging constraints on A/C system performance in the early engineering stage of any program.

When an assembly plant finishes building a vehicle, the entire fuel system is dry. Before the first engine start, a few gallons of fuel are added to the tank. On this very first “green start,” air must be purged out of the entire fuel system. This is done by running the tank fuel pump while opening one or more fuel injectors to push air out. There are special green engine calibrations in the ECM that govern this process. If this purge process is too long, fuel will come out of the injectors and flood the cylinders, making a start impossible for several minutes or in extreme cases, causing the engine to hydro-lock. If the purge process is too short, and air remains in the fuel system, the vehicle will crank for an extended time, and then stall soon after the engine first fires. Both of these events cause significant disruptions for the plant.

There is an opportunity to improve the process by which these calibration values are set. Recently, engineers have had to travel to assembly plants to develop their calibrations on saleable vehicles, and then push post-RPR calibrations to the plant. The goal of this project is to develop a calibration process and tool that would allow engineers to obtain reliable green engine calibrations with the least time and resource investment possible. The resultant calibration tool utilizes the correlation between existing, functional calibrations from across the GM portfolio and the basic vehicle parameters of engine displacement and vehicle width. From the tool, a calibrator receives working values for 35 green engine calibrations and guidance on how to test this final calibration in vehicle. The tool was validated on MY2020 vehicles with L3B and LFV engines. This new calibration tool and method will be presented to the Open Loop Fuel Control TRB. The new method can be utilized across all direct-injection engines. Cost savings will be realized in terms of lessening disruptions to assembly plants, minimizing calibrator travel to plants, and eliminating the need to push post-RPR calibrations to solve green engine problems.

General Motors recently released its CarbonPro pickup box, the first carbon fiber reinforced thermoplastic pickup box in the industry. As with most composite applications, scrap material is a large cost driver. One of the benefits of being a thermoplastic is that the scrap material can be recycled. The team was able to identify 3 accessory applications that General Motors would benefit from implementing the recycled material. An estimate from the supplier showed a 35% cost savings could be gained by switching from the current material to recycled carbon fiber reinforced plastic on an existing accessory. The team is proceeding with optimizing the designs of the 3 accessories chosen, and if we are able to achieve the same cost savings margin and utilize all of the scrap material we would save General Motors over $1.3 million/year. Along with the cost savings this would further expand the usage of composite materials and can be used to further promote the CarbonPro package. Furthermore, the recycled material will also provide a weight savings for the selected accessories.

The Chassis Controls Hardware in the Loop lab at General Motors plays a key role in chassis controls software and hardware development. A significant portion of the team’s workload consists of testing value-added features, so a tool to post-process these results is desired. A tool was developed in Matlab and tested by engineers in the chassis controls lab. Based on time saved, it is estimated that this tool will save $79,200 annually. In addition to monetary savings, this tool will provide more efficient communication between GM and suppliers and help streamline the development process for the Hardware in the Loop lab.

Vehicle level performance requirements were revisited for level 4 and 5 autonomous systems that automatically accelerate and brake with the flow of traffic on expressways. The first is providing a methodology for peak braking capability of an autonomous vehicle from an expressway cruising speed. The second is providing a methodology for vehicle jerk requirement at expressway cruising speeds to ensure driver comfort. For this report, objective measurements were found using a model which compared the following distance, speeds, and deceleration profiles between a leading and following vehicle in an emergency braking scenario, and subjective evaluations were performed in a Cadillac CT6 with SuperCruise and Full Speed Range Adaptive Cruise Control. If too high a maximum vehicle deceleration capability requirement is chosen, optimizations in other trade off areas such as tire durability could be underspecified. If too low-performance maximum vehicle deceleration capability is chosen, vehicle could be unsafe for the various expressway driving scenarios an autonomous vehicle would encounter, as well as vehicles may cut in front of the vehicle on the expressway compromising ride and violating safety requirements. To optimize maximum vehicle deceleration capability while providing a safe and smooth ride to passengers, requirements must set constraints for the design of various subsystems in both hardware and software. Subjective and objective measurements were combined into a methodology to calculate a Minimum Safety TTC based on maximum vehicle deceleration, and a comfort jerk requirement.

During the development of a vehicle, validation of subsystems following a defined set of requirements is important to ensure vehicle is safe, reliable and meeting customer life span usage. In order to meet the target with high confidence level but within a certain period of time that does not increase the development time, a significant amount of properties are necessary. The purpose of this project was to determine a plan to reduce the physical properties required to validate a vehicle design, by improving analytical tool inputs and robustness supported by process enhancements.

Based on the results of discussions during the project with the subject experts, a property reduction strategy was generated that achieved a 40% reduction in full vehicle bucks, and over 62% reduction in door sets required for validation testing. For the purposes of this report, and to protect confidential company information, it can be assumed that this reduction is proportional to the costs saved by this reduction.

The automation was implemented for a large-scale manufacturing plant located in Saline, MI. Due to the plant’s size, handling such data and interpreting requires intensive work in a given time frame. Namely, the company was interested in calculating the efficiency of the individual assembly lines at the site when my internship started. During my internship, I developed an automation script to extract data from a company network in a matter minutes, enabling the engineers at the site to use real-time data.

Before this automation took place, the company did not have a resource and manpower to calculate the efficiency of the lines daily. However, with the automation this became possible. Further with the use of this calculations, the engineering team was able to track the productivity and continuously improve the assembly lines. Overall, the company reduced time in calculating the efficiency by hand and saved cost by running an efficient plant.