Get in touch: 901-734-4594 | jcalebgood@gmail.com
Skip to Content

Caleb Goodwin

• Introduction
• Vita
• Publications

CATEGORIES

• Research
• Publications
• Vita
• About Me

EMERGENCY CARE IS A COMPLEX ADAPTIVE SYSTEM

My current research views medical error as an emergent property of the complexity of emergency care. Once understood, the systemic behavioral states will function like physiological identifiers for biological organisms and as such can provide queues that permit identification of potential failures before death of the system occurs.

SEMANTIC WEB

Given a network of thousands of wired and wireless heterogeneous sensors, how can the sensors be discovered and dynamically utilized for applications that might not have been anticipated on initial deployment?

BIOMEDICAL DATA VISUALIZATION

The focus of this research was the development of an interactive defibrillation simulation known as DefibViz which utilizes data visualization techniques to allow for exploration of voltage gradients throughout the human torso.

DATA VISUALIZATION OUTREACH

The focus of this research was the interdisciplinary development of data visualization tools for pedagocial use and high-school outreach.

Semantic Web Research Overview

August 2005-August 2007

Project Overview

The Semantic Web research for intelligent sensors was conducted under the supervision of Dr. David J. Russomanno at The University of Memphis Department of Electrical and Computer Engineeringwith the Center for Advanced Sensors.

The topic that I investigated during my research was the following: Given a network of thousands of wired and wireless heterogeneous sensors, how can the sensors be discovered and dynamically utilized for applications that might not have been anticipated on initial deployment? I inherited the work-in-progress ontology called OntoSensor which is a knowledge model describing sensor properties. A laboratory test bed was created using wireless sensors that referenced OntoSensor.

The next phase of my research can be summarized as a search engine leveraging a service-oriented architecture that utilizes an ontology instead of statistical methodology to discover live sensor services. The architecture leveraged the Universal Description Discovery and Integration (UDDI), which is a industry standard syntax-based search registry for Web services. The developed architecture is composed of two primary modules: i) UDDI and ii) semantic sensor matchmaker. The semantic sensor matchmaker works in conjunction with UDDI to discover sensor services using semantic processing. Ontological knowledge pertinent to the satisfying a query is updated/inserted into the UDDI database. The process is analogous to using a set of compiled heuristics first (that is, the explicit facts in the UDDI database) then falling back on deeper knowledge (the sensor ontology) if the heuristics do not provide an acceptable solution. The research work resulted in one conference paper which I presented at the 2007 International Conference on Semantic Web and Web Services.

Sensor Ontology

Project Overview

OntoSensor is a Semantic Web compatible ontology developed using Protege. OntoSensor references and extends the IEEE Suggested Upper Merged Ontology (SUMO), which defines general concepts and associations. OntoSensor is based in-part upon SensorML, which defines associations and properties common to sensors. OntoSensor deviates from SensorML since it lacks the semantic richness, such as axiomatic-grounded terms, which may be required for automated data fusion and inference in a distributed sensing environment.

Ontology-Based Sensor Network Prototype

Project Overview

The ontology-based sensor network prototype is comprised of two computers that serve as base stations and store the data collected from the MIB510 and MIB520 network gateways. The base stations register their respective services with the sensor service broker developed for my thesis and published in 1 and 2. The sensor service broker allows semantic discovery of the registered services. Each base station receives data from a sensor network. The wireless sensors form ad-hoc communication links to route the collected data back to the base station. The base stations run Crossbows MoteView application to retrieve data from the network gateways and the data is stored in a PostGRE database. In addition, each base station executes custom software developed for during my thesis that extracts sensor data and metadata into OWL repositories that reference OntoSensor. The software is a preliminary implementation of a Web service that is evoked at the base stations for selectively storing sensor data into OWL repositories within the networked environment.

The base stations also serve as a logical interface for tasking the individual sensors through Web services. Base stations also provide physical storage aggregation nodes for sensor data and metadata that can be used in response to queries by agents, rather than involving the individual sensor nodes in responding directly to queries.

Semantic Web Service Discovery

Project Overview

This architecture includes the Prolog implementation of a UDDI inspired repository, sensor matchmaker, and sensor service interface. The sensor matchmaker is registered in the Prolog UDDI registry and is invoked by the sensor service interface to provide semantic search capabilities. The matchmaker stores in UDDI the semantic metadata pertinent for satisfying a query so that subsequent syntactic search queries can be satisfied directly from the UDDI repository and avoid the overhead of semantic-based search. The sensor matchmaker has the following capabilities; i) search via generalization; ii) search via specialization; iii) performance property search; iv) supported applications search; and v) loading additional semantic metadata to satisfy queries.

Relevant Publications

D.J. Russomanno and J.C. Goodwin (2008) OntoSensor: An Ontology for Sensor Network Application Development, Deployment, and Management, Handbook of Wireless Mesh and Sensor Networking, McGraw Hill

J.C. Goodwin (Thesis) Ontology Integration within a Service-Oriented Architecture for Sensor Networks.

J.C. Goodwin and D.J. Russomanno. (In Peer Review) Ontology Integration within a Service-Oriented Architecture for Expert System Applications using Sensor Networks, Journal of Expert Systems

J.C. Goodwin and D.J. Russomanno. (2007) Survey of Semantic Extensions to UDDI: Implications for Sensor Services, The 2007 International Conference on Semantic Web and Web Services

J.C. Goodwin and D.J. Russomanno. (2006, Poster) An Ontology-Based Sensor Network Prototype Environment, Fifth International Conference on Information Processing in Sensor Networks, Nashville, TN.

THESIS

Back

Ontology Integration within a Service-Oriented Architecture for Sensor Networks

This thesis describes the development and implementation of an ontology-based network centric service-oriented architecture for discovery of sensor services using semantics. The architecture enables software agents to discover ubiquitous sensors referencing machine-interpretable service descriptions, such as those proposed by the Semantic Web effort, enabling on-the-fly utilization of sensors for applications that might not have been anticipated on initial deployment. The developed architecture overcomes some of the limitations of the current service-oriented architecture technologies that rely solely on XML data structures and syntax-based search mechanisms for discovery of Web services. The architecture developed in this thesis allows for the generalization, specialization, performance property, supported application, and instance semantic processing to satisfy queries. Furthermore, the architecture seeks to enhance the syntactic search for sensor services by extracting semantic metadata from a sensor ontology pertinent to satisfying a query and saving it in the service registry.

Journal

Back

Ontology Integration within a Service-Oriented Architecture for Expert System Applications using Sensor Networks

This paper describes the development of an architecture for the discovery of sensor services leveraging ontology-based semantics in the search query. A prototype has been implemented based upon the architecture and can be used to support the development of expert system applications in which sensors of certain types, operational capabilities, or physical properties are required to support applications within a network-centric environment. In the prototype, sensor services are listed in a registry that references a machine-interpretable ontology. The registry conforms to the Universal Discovery and Description Interface (UDDI) specification, but it is augmented with semantic matching via an ontology to increase the likelihood that relevant sensor services are discovered when needed by expert system applications.

Conference

Back

Survey of Semantic Extensions to UDDI: Implications for Sensor Services

Abstract-The ability for software agents to discover, query, and task ubiquitous sensors requires machineinterpretable service descriptions, such as those proposed by the Semantic Web effort. Descriptions that support deep semantics will enable on-the-fly utilization of sensors for applications that might not have been anticipated on initial deployment. Semantic Web service discovery and dynamic composition requires formal semantic descriptions of inputs, outputs, preconditions, and effects of services. Universal Description Discovery and Integration (UDDI) provides a registry for publication and discovery of Web services, but it lacks the semantics needed for discovery and interoperation as envisioned by the Semantic Web community due to UDDI’s syntax-based search. This paper surveys representative approaches for incorporating semantic capabilities within the existing UDDI infrastructure and then proposes an architecture for sensor services within an ontology-based networkcentric environment.

Book Chapter

Back

OntoSensor: An Ontology for Sensor Network Application Development, Deployment, and Management

This chapter describes the potential utilization of an ontology in heterogeneous sensor network environments to facilitate discovery, query, tasking, inference, and interoperation of a myriad of sensor types. As sensor networks advance, they will no longer be dedicated to specific applications in which a priori knowledge of the sensors’ capabilities and access methods are procedural bundled within application-specific software. In dynamic scenarios, declarative knowledge sources are required to support the on-the-fly utilization of sensors by software agents. A laboratory environment and evolving sensor ontology have been constructed to experiment with such scenarios. OntoSensor has been implemented using the Web Ontology Language to model sensor properties, associations, and services providing meta data about sensor types, as well as the knowledge required for subsequent interoperability given that the sensors may contribute to an agent’s overall goal.

Poster

Back

Ontology-based Sensor Network Prototype Environment

Abstract—This paper describes work-in-progress development of an ontology-based, sensor network prototype environment to facilitate research in distributed, heterogeneous sensor inference, fusion, tasking and control. Currently, the prototype environment consists of wireless sensing capabilities that include temperature, acceleration, GPS, light, barometric pressure, magnetic field and acoustic measurements, with wired visible, infrared, and other high-bitrate sensors pending integration. Communication links formed by the sensors permit the aggregation of data at the base stations. Each base station includes a process that generates a sensor data repository from the raw percepts, which is marked up using the Web Ontology Language (OWL) to reference the OntoSensor ontology. Each sensor type in OntoSensor is defined using concepts, associations, and services providing meta data about each sensor, as well as requisite knowledge for interoperability and data fusion. A software agent developed in SWI PROLOG loads the OWL sensor repositories into its knowledge base for subsequent application. Currently, the agent only supports ad-hoc queries of the sensor repositories to discover trends in the measurements. Plans for future work include demonstration of proof-of-concept utility of sensor ontologies in distributed sensor processing and control.

© by Caleb Goodwin 2008