Ali Akgunduz

ADVICE: A virtual environment for Engineering Change Management

In this paper, we analyze Engineering Changes (EC), which are modifications in forms, fits, functions, materials, or dimensions in components constituting the design. We propose a new approach for processing these changes within a Virtual Collaborative Design Environment. This environment, named Active Distributed Virtual Change Environment (ADVICE) offers an Engineering Change Management solution by merging graphical and parametric data involved in the process into a virtual platform, which improves the comprehension of users and hence decreases the time required for review. ADVICE provides smart user support for predicting Engineering Changes to be triggered due to a specific change and for offering priorities to Engineering Change Requests. This is managed by employing Sequential Pattern Mining techniques to process captured Engineering Change history with Prioritization and Change Propagation mechanisms. For verifying these mechanisms, experiments involving synthetic data are conducted. The experiments present the capability of ADVICE to facilitate Engineering Change Management.

A comparative study of heuristic algorithms on Economic Lot Scheduling Problem

The Economic Lot Scheduling Problem (ELSP) has been well-researched for more than 40 years. As the ELSP has been generally seen as NP-hard, researchers have focused on the development of efficient heuristic approaches. In this paper, we consider the time-varying lot size approach to solve the ELSP. A computational study of the existing solution algorithms, Dobsons heuristic, Hybrid Genetic algorithm, Neighborhood Search heuristics, Tabu Search and the newly proposed Simulated Annealing algorithm are presented. The reviewed methods are first tested on two well-known problems, those of Bombergers [Bomberger, E. E. (1966). A dynamic programming approach to a lot size scheduling problem. Management Science 12, 778-784] and Mallyas [Mallya, R (1992). Multi-product scheduling on a single machine: A case study. OMEGA: International Journal of Management Science 20, 529-534] problems. We show the Simulated Annealing algorithm finds the best known solution to these problems. A similar comparison study is performed on various problem sets previously suggested in the literature. The results show that the Simulated Annealing algorithm outperforms Dobsons heuristic, Hybrid Genetic algorithm and Neighborhood search heuristics on these problem sets. The Simulated Annealing algorithm also shows faster convergence than the best known Tabu Search algorithm, yet results in solutions of a similar quality. Finally, we report the results of the design of experiment study which compares the robustness of the mentioned meta-heuristic techniques.

A tabu search algorithm for solving economic lot scheduling problem

The economic lot scheduling problem has driven considerable amount of research. The problem is NP-hard and recent research is focused on finding heuristic solutions rather than searching for optimal solutions. This paper introduces a heuristic method using a tabu search algorithm to solve the economic lot scheduling problem. Diversification and intensification schemes are employed to improve the efficiency of the proposed Tabu search algorithm. Experimental design is conducted to determine the best operating parameters for the Tabu search. Results show that the tabu search algorithm proposed in this paper outperforms two well known benchmark algorithms.

A Linear Programming Solution for Exact Collision Detection

This paper addresses the issue of real-time collision detection between pairs of convex polyhedral objects undergoing fast rotational and translational motions. Accurate contact information between objects in virtual reality based simulations such as product design, assembly analysis, performance testing and ergonomic analysis of products are critical factors to explore when desired realism is to be achieved. For this purpose, fast, accurate and robust collision detection algorithms are required. The method described in the text models the exact collision detection problem between convex objects as a linear program. One of the strengths of the proposed methodology is its capability of addressing high speed interframe collision. In addition to the interframe collision detection, experimental data demonstrate that mathematical programming approaches offer promising results in terms of speed and robustness as well. DOI: 10.1115/1.1846053

Tracking-based depth recovery for virtual reality applications

Abstract.This paper describes a technique for tracking-based reconstruction from camera images. The position and orientation data are provided by a commercial laser tracker rigidly attached to the camera. The tracker enables the determination of the extrinsic camera parameters. Focal length is automatically calibrated in parallel with the depth computation process by means of a Kalman filter. In addition to its simplicity, our technique enables the fast generation of three-dimensional models from two-dimensional images to be used in virtual reality applications.

Two-step 3-dimensional sketching tool for new product development

This paper discusses a two-step virtual reality based conceptual design tool that enables industrial designers to create sketches of their ideas in 3-dimensional space in real time. In the developed sketching tool, the rough shapes of products are generated by tracing the trajectory of the data-gloves worn by the designer. In the model a practical solution is provided to reduce the generation of unnecessary control points. This is achieved by representing each control point by a spherical volume. Once the rough sketching is completed, NURBS surfaces are constructed by the limited number of reference points that are selected from the initial sketch by using a virtual pen. The two-steps sketching technique enables designers to perform their artistic characteristics freely in an intuitive environment and also enables designers to generate parametric representations of the surfaces to be used in CAD/CAM systems for further analysis.

An airline revenue management pricing game with seat allocation

This paper studies a horizontal fare-pricing competition between two airlines having a single flight leg. Two distinct scenarios are considered. First, the two airlines price competition for the pre-committed booking limits is analysed. The problem is studied under deterministic price sensitive demands. The existence of unique pricing strategies at Nash equilibrium is shown. In the second scenario, a joint seat allocation and fare-pricing competition model for stochastic demand is proposed. A numerical analysis is presented to demonstrate the impacts of various market conditions on the payoffs, booking limits and pricing strategies of the competing airlines.

Automated end-user behaviour assessment tool for remote product and system testing

This paper introduces a virtual reality (VR) based automated end-user behavioural assessment tool required during product and system development process. In the proposed method, while VR is used as a means of interactive system prototype; data collection and analysis are handled by an event based system which is adopted from human-computer interaction and data mining literature. The overall objective of the study is developing an intelligent support system where physical prototyping and the needs for human involvement as instructors, data collectors and analysts during the user assessment studies are eliminated. The proposed method is tested on a product design example using 27 subjects. The experimental results clearly demonstrate the effectiveness of the proposed user assessment tool.

Pair-Wise Preference Comparisons Using Alpha-Peak Frequencies

Understanding of user needs and preferences, and incorporating them into product designs are crucial for new products to be successful in todays highly competitive marketplace. This paper introduces a unique methodology to capture user preferences from biological signals. The proposed method successfully utilizes one of the most complex phenomenons of the human body, bio-signals for preference extraction. Bio-signals are magnetic fields that are generated by the neurons as a response to a persons surroundings. Due to its complexity, interpretation of bio-signals in the form of natural communication channels is extremely difficult. On the other hand, our experiments with potential users of various types of products revealed that detectable shifts from steady state levels of brain signals in the positive or negative direction are strongly correlated with the individuals preferences on alternative designs. From experiments conducted on 14 different subjects we conclude that when the alpha-peak frequencies EEG signals at the 8--12 Hz band are captured during a users interaction with two competing product design solutions, the lower alpha-peak value indicates a higher preference over the competing product for right-sided subjects. The opposite effect is observed for the left-sided subjects.

An Enhanced Optimization Model for Scheduling Freight Trains

We propose a new dynamic row/column management algorithm for the schedule of freight trains in a single/double track railway system. While many works have already been devoted to train scheduling, previously published optimization models all suffer from scalability issues. Moreover, very few of them consider the number of alternate tracks in the sidings for train meets, as well as the delay incurred by trains that take sidings. We propose a non time-indexed model, which takes into account such constraints, and we design a solution scheme with iterative additions/removals of constraints/variables in order to remain with a manageable sized mathematical program, while still ensuring convergence to an optimal solution. Numerical results are presented on data instances of Canada Pacific Railway. We evaluate the performance of the optimization model, and the sensitivity of the train schedules to the length of the operation hours, the length of the trains, and the departure times.Copyright