Sundar Krishnamurty

Ontologies for supporting engineering design optimization

This paper presents an optimization ontology and its implementation into a prototype computational knowledge base tool dubbed ONTOP (ONT ology for OP timization). Salient features of ONTOP include a knowledge base which incorporates both standardized optimization terminology, formal method definitions, and often unrecorded optimization details, such as any idealizations and assumptions that may be made when creating an optimization model, as well as the model developer’s rationale and justification behind these idealizations and assumptions. ONTOP was developed using Protege, a Java-based, free open-source ontology development environment created by Stanford University. Two engineering design optimization case studies are presented. The first case study consists of the optimization of a structural beam element and demonstrates ONTOP ’s ability to address the variations in an optimal solution that may arise when different techniques and approaches are used. A second case study, a more complex design problem which deals with the optimization of an impeller of a pediatric left ventricular heart assist device, demonstrates the wealth of knowledge ONTOP is able to capture. Together, these test beds help illustrate the potential value of an ontology in representing application-specific knowledge while facilitating both the sharing and exchanging of this knowledge in engineering design optimization.© 2006 ASME

A robust multi-criteria optimization approach

The overall value of a product is typically determined by its performance with respect to multiple measures. As such, the product design task will be simplified if all these performance measures were optimized simultaneously. Another significant factor that determines product quality is its sensitivity to external or uncontrollable variations. To effectively address these issues in product design, this paper presents a novel robust multiple criteria optimization (RMCO) approach that integrates multi-objective optimization concepts with statistical robust design techniques. In this approach, Pareto-optimal robust design solution sets are obtained with the aid of design of experiment set-ups, that utilize ANOVA results to quantify relative dominance and significance of design factors. Application of this method to engineering design problems is illustrated with the aid of two case studies including a mechanism dimensional synthesis problem.

A generalized exact gradient method for mechanism synthesis

Currently no single method, analytical or numerical, exists that is well suited for the dimensional synthesis of general mechanisms. One of the main reasons that even the state-of-the-art optimization methods are not utilized for developing a generalized mechanism synthesis methodology is due to the difficulties in obtaining the partial derivatives necessary for optimization. This paper presents the development of a new exact gradient method that offers an efficient, reliable, and accurate solution procedure for the optimal synthesis of generic mechanisms. The application of this exact gradient method is illustrated with the aid of practical six-bar and eight-bar mechanism examples.

Optimal synthesis of mechanisms using nonlinear goal programming techniques

Abstract The application of multiple objective optimization techniques based on the methods of nonlinear goal programming to perform optimal synthesis of general planar mechanisms is presented. In this optimization technique, the objectives of the design problem are first identified and prioritized according to their relative importance. The mechanism analysis is then performed to identify the design variables and their relationships to the dependent variables. The nonlinear goal programming technique is employed to determine the optimal values for the design variables that best satisfy the desired objectives of the problem. The uniqueness of this optimization technique lies in its ability to include all the objectives directly in the optimization process. This method, thus, eliminates the need to identify or combine the objectives into a single objective for the purpose of optimization. Several illustrative mechanism synthesis problems with multiple objectives were studied for optimal solutions using the nonlinear goal programming technique and the method produced accurate and acceptable solutions in all cases. The application of this method to three typical industrial type mechanism synthesis problems is presented and the results are discussed.

Design and realization of a portable data logger for physiological sensing [GSR]

A microcontroller-based data logger has been designed, prototyped, and field-tested for recording galvanic skin response data and relaying them to a computer for physiological analysis. Focusing on system design issues concerning battery-driven ambulatory applications, this paper presents a special data compression algorithm based on relative encoding to optimize memory utilization and reduce data transfer time. Data flow coordination and timing control are enabled by a PIC microcontroller. A handheld prototype measuring 180/spl times/90/spl times/30 mm was built and tested for ease of use, safety, and reliability.

A Web-Based Environment for Documentation and Sharing of Engineering Design Knowledge

This paper presents the foundation for a collaborative Web-based environment for improving communication by formally defining a platform for documentation and sharing of engineering design knowledge throughout the entire design process. In this work an ontological structure is utilized to concisely define a set of individual engineering concepts. This set of modular ontologies link together to create a flexible, yet consistent, product development knowledge-base. The resulting infrastructure uniquely enables the information stored within the knowledge-base to be readily inspectable and computable, thus allowing for design tools that reason on the information to assist designers and automate design processes. A case study of the structural optimization of a transfer plate for an aerospace circuit breaker is presented to demonstrate implementation and usefulness of the knowledge framework. The results indicate that the ontological knowledge-base can be used to prompt engineers to document important product development information, increase understanding of the design process, provide a means to intuitively retrieve information, and seamlessly access distributed information.© 2008 ASME

Design and Innovative Methodologies in a Semantic Framework

Significant expenditure and effort is devoted to the never ending search for reduced product development lifecycle time and increased efficiency. The development of Semantic Web technologies promises a future where knowledge interchange is done seamlessly in open distributed environments. This paper illustrates how Semantic Web technologies in their current state of development can be effectively used to deploy an infrastructure supporting innovation principles and the engineering design processes. A mechanical design was chosen to model the initial phase of a design project using semantic ontologies. This included a set of design requirements, creating a functional model, and making use of the Theory of Inventive Problem Solving (TIPS). The ontology development strategy is built on a combination of larger domain knowledge ontologies and simple process ontologies. Linked user requirements, engineering design, and functional modeling ontologies facilitated the application of TIPS through a set of semantic rules to generate design recommendations. The developed semantic knowledge structure exemplifies a practical implementation of a functional model which served as a record of the design process and as a platform from which to gain additional usefulness out of the stored information.Copyright

Semantic methods supporting engineering design innovation

In this paper, we present a metric based on semantic relatedness which operates on semantic knowledge representations of engineering design and show how it can support design innovation. Our semantic knowledge representation is composed of an ontology representing design concepts using the National Institute of Standards and Technology (NIST) functional basis formalism. We assert that the uniqueness of a design concept is directly proportional to the mean semantic distance between itself and the set of competing design concepts represented as instances within our functional basis ontology. This leads to our Semantic Relatedness Uniqueness Metric called SeRUM. SeRUM draws upon semantic functional model representations of design concepts and computer science semantic relatedness techniques. SeRUM provides design teams a measure of their effectiveness in terms of generating unique design concepts. To highlight SeRUM’s application in engineering design innovation, a design innovation case study is detailed and the results are discussed.

Exploring the Safety and Therapeutic Effects of Deep Pressure Stimulation Using a Weighted Blanket

Abstract This paper presents the results of a concurrent, nested, mixed methods exploratory study on the safety and effectiveness of the use of a 30 lb weighted blanket with a convenience sample of 32 adults. Safety is investigated measuring blood pressure, pulse rate, and pulse oximetry, and effectiveness by electrodermal activity (EDA), the State Trait Anxiety Inventory-10 and an exit survey. The results reveal that the use of the 30 lb weighted blanket, in the lying down position, is safe as evidenced by the vital sign metrics. Data obtained on effectiveness reveal 33% demonstrated lowering in EDA when using the weighted blanket, 63% reported lower anxiety after use, and 78% preferred the weighted blanket as a calming modality. The results of this study will be used to form the basis for subsequent research on the therapeutic influence of the weighted blanket with adults during an acute inpatient mental health admission.

A Semantic Information Model for Capturing and Communicating Design Decisions

A semantic information model to improve reuse and communication of engineering design knowledge is presented in this paper. We consider design to be a process involving a sequence of decisions informed by the current state of information. As such, the information model developed is structured to reflect the conceptualizations of engineering design decisions with a particular emphasis on semantically capturing design rationale. Through the approach presented, knowledge reuse is achieved by communicating design rationale. A case study is presented to illustrate two key features of the approach: (1) seamless integration of separate modular domain ontologies and instance knowledge related to engineering design that are needed to support decision making and (2) the explicit documentation of design rationale through design decisions.