- Master Projects
- Timing analysis for ADAS Architectures
- Print head optimization; exploring trade-offs between quality and productivity
- Modeling of Oce production printers as manufacturing systems
- Modeling component tolerances for paper path designs
- Performance Optimization by Playing Games
- Modeling and analysis of the workflow around wide format roll-to-roll Oce printers
- Visualization and analysis of data caching
- Multi-core Datapath Contention Modelling
- Interfacing CompSOC platform with Advanced Driver Assistance Systems Testbed
- Visual Simultaneous Localization and Mapping using Focal-place Processors
- Productivity printer performance modeling
- Calculation of stacking sequences
This page contains previous topics of bachelor or master projects. Suggestions for new project ideas can always be discussed.
Timing analysis for ADAS Architectures
TNO is the world leader in truck platooning technology. This technology allows a group of trucks to become linked to each other, using data transmitted over radio communication, forming a virtual “vehicle train” driven by the front vehicle driver only (the other drivers become passengers).
The technology that enables truck platooning is called Cooperative Adaptive Cruise Control (CACC). To guarantee correct CACC it is important to accurately derive the data transmission latencies between trucks. The latency calculations should take into account not only the radio communication delays, but also the delays introduced by the data flowing from one component of the truck hardware platform to another. In this assignment, you will contribute to development of a systematic methodology to ascertain the worst-case-scenario end-to-end latencies together with its statistics, in a way that is modular and captures the shared resources interactions.
The assignment will give you the opportunity to get acquainted with real-life ADAS architectures, work in a dynamic environment on automated driving functions at TNO Helmond and to link the practical and theoretical aspects of timing analysis of ADAS architectures.
Project description: Timing analysis for ADAS Architectures
Print head optimization; exploring trade-offs between quality and productivity
We want to investigate the impact of reconfiguration of the print head of a printer on productivity and quality. Stringent productivity targets force engineers to look at ways to improve productivity, without losing (too much) product quality. Reconfigurations during operation generally improve quality, but cost time and hence, productivity. On-line reconfiguration of a print head can sometimes be avoided by tolerating small deviations from the optimal configuration within an acceptable working range for a particular product.
In this assignment you will model the quality versus productivity trade-off that emerges from the scheduling freedom of the print head range. The first step is to model the quality and productivity depending on reconfiguration strategies, and define and study the trade-off space. The second step is to find the optimal trade-offs and scheduling strategies that lead to these trade-offs, for given specification ranges and sequences of sheets and their media types.
Project description: Print head optimization
Modeling of Oce production printers as manufacturing systems
In this assignment we look at modeling the paper path of a large-scale production printer as a manufacturing system and optimizing the sheet schedules. Such a printer prints thousands of duplex images per day (e.g. books for Amazon), transaction printing (bank statements) on different media. The paper path of the printer is large enough to fit up to 100 sheets at the same time. As these sheets are to be printed on the front and the back by the same print head, the sheets printed once should be efficiently interleaved with unprinted sheets. The media and size of the paper may influence the minimum time between subsequent sheets at the print head, and the return loop needs to operate within a minimum and maximum speed.
To investigate the large-scale production printer in terms of timing and productivity, we propose to use a mathematical model that allows automated analysis and optimization of the manufacturing system. Recently, we have developed a formal modeling approach to model manufacturing systems. In this assignment, you will investigate the application of this approach to create a behavioral model of the large-scale production printer. This model needs to capture the timing aspects, and in which order, and when, the printer can execute certain tasks. This model can then be used to optimize the printing sequence. If certain aspects of the production printer cannot be captured using the current modeling approach, or cannot be analyzed with existing analysis methods, the approach needs to be extended.
Project description: Modeling of Oce production printers as manufacturing systems
Modeling component tolerances for paper path designs
The aim of this project is to include the tolerances of (transportation) components such as pinches, switches and belts into an existing optimization model. The model optimizes the paper path lengths and timing profiles for particular sequences of sheets through a direct transcription model. In this assignment you will extend the model by taking into account imperfections, and create a model that propagates the positional errors/skew to optimize timing profiles for sheets.
Project description: Modeling component tolerances for paper path designs
Performance Optimization by Playing Games
In this assignment you will look at performance optimization of production systems using game theory, the mathematical theory that deals with modeling of conflict and cooperation between intelligent rational decision-makers. Game theory is an active research area that has applications in many fields, such as economics, logic, psychology, computer science, and electrical engineer- ing. Game theory can also be used to optimize the performance of systems in an unpredictable or uncontrollable environment. In this case, the system controller is modeled as a player, and the environment is modeled as its opponent player. Synthesizing an optimal controller corresponds to finding a winning strategy for the controller player in terms of the game. There are games that can optimize for throughput, or try to prevent that unsafe system states are ever reached.
In this assignment, you will investigate existing games, and develop new games if needed, that can be used for performance analysis and optimization of systems. Because systems are often optimized for multiple criteria, and need to adhere to certain requirements, multi-objective optimization under constraints is part of the assignment. For instance, finding all controllers that satisfy a throughput bound, and then find the best one in terms of minimal latency and maximal energy efficiency.
Project description: Performance Optimization by Playing Games
Modeling and analysis of the workflow around wide format roll-to-roll Oce printers
In this assignment we look at modeling and analysis of the workflow around wide format roll- to-roll printers. Such a printer prints large images (posters, banners, etc) on different media. The operators have to change rolls (when a roll is empty or another media type is needed), to fill ink (ink handling can be done during print), and remove waste. Furthermore, the operator may have to perform preprocessing and offline postprocessing (for example laminating) on specific jobs. To some extent, the operator can reorder print jobs. To investigate the workflow in terms of timing and productivity, we propose to use a mathematical model that allows automated analysis and optimization of the workflow.
Project description: Modeling and analysis of the workflow around wide format roll-to-roll Oce printers
Visualization and analysis of data caching
Many embedded systems are data intensive. Examples include the video processing system in an interventional X-ray machine, and the processing of image data in the datapath of high-end printers. Often, these systems are implemented, at least partially, in software which runs on multi-core embedded platforms. Furthermore, there are strict performance requirements, e.g., on the throughput of the system. Clearly, the software implementation has significant impact on the system performance. One key aspect in these data-intensive systems is how the software implementation uses the hardware data caches. An actual assessment whether cache usage can be optimized and what its system-level performance benefits are, however, is difficult because it requires the user to gather detailed information from several sources through specialized tools (e.g., Intel VTune), and to manually combine information for the final interpretation. The goal of this assignment is to provide an integrated approach to the above-sketched problem based on execution traces of the system.
Project description: Visualization and analysis of data caching
Multi-core Datapath Contention Modelling
The datapath is a central part of printers and copiers and is responsible for the image processing that is necessary to produce output of required quality. A key performance indicator of the datapath is its performance, typically expressed as throughput in processed image-data per minute. The operations in the datapath are data-intensive and are often implemented in software on multi-core embedded platforms. The programming can thus have a significant effect on the performance through its use of the data caches. Currently performance offered by platform components rapidly increases. The graduation assignment is to analyze and optimize the image processing in print datapaths for multi-core platforms shared with other software components, under the constraint that the image processing performance may not decrease, in order to guarantee sufficient throughput for the printer.
Project description: Multi-core Datapath Contention Modelling
Interfacing CompSOC platform with Advanced Driver Assistance Systems Testbed
EMC2 – “Embedded Multi-Core systems for Mixed Criticality applications in dynamic and changeable real- time environments” is an ARTEMIS Joint Undertaking project started in 2013. The objective of EMC2 is to establish Multi-Core technology in all relevant Embedded Systems domains. Within the EMC 2 project, one domain of interest is Advanced Driver Assistance Systems (ADAS) for next generation vehicles. To this end, a testbed for ADAS is being developed by Technolution (www.technolution.eu). The idea of this master project is to build up an interface between ADAS testbed and CompSOC platform from TU/e and investigate various architectural and algorithmic design aspects.
Project description: Interfacing CompSOC platform with Advanced Driver Assistance Systems Testbed
Visual Simultaneous Localization and Mapping using Focal-place Processors
Assignment: analysis, simulation, and coding of common algorithms used for monocular visual SLAM. You will carry out an evaluation to choose the best algorithms suited for FP. The next step entails FP parallelization of the results for FP processors.
This project will be carried out at Technology Research Center, Finland.
Project description: Visual Simultaneous Localization and Mapping using Focal-place Processors
Productivity printer performance modeling
Assignment: Create and enhance models for predicting the performance of the production printers of Canon Océ. For all kinds of print shop optimization, such as workload planning, it is necessary to have a fast and accurate estimation of how much time a print job will need to execute. This problem is quite difficult due to the complex layout of the printer and the non-trivial scheduling problem.
Project description: Productivity printer performance modeling
Calculation of stacking sequences
Assignment: Vanderlande provides automated material handling systems and services that focus on improving customers’ logistics processes and increasing their logistics performance throughout the entire life cycle. Automated Case Picking is one of such systems.Enhance the algorithm for calculating how to stack pallets with products, so that the resulting stack is stable and efficiently packed.
See Automated Case Picking for an example of Vanderlande solutions.
Project description: Calculation of stacking sequences