Dr. Ken Holladay of Southwest Research Institute

Biography:

Dr. Holladay has extensive experience in the areas of machine learning, software architecture, system engineering, and human computer interaction (HCI). In addition to project work, Dr. Holladay has also been the principal investigator (PI) for several Southwest Research Institute (SwRI) internal research and development (IR&D) projects in these areas, including a Staff Renewal project to develop and teach a graduate level Data Science course for the Computer Science department at UTSA.

In the machine learning domain, he led an IR&D program that analyzed a large volume of real-time turbofan engine data collected during thousands of flight hours. This effort successfully developed an anomaly detection framework that identified rare compressor stall events with very low false positive and false negative rates, even when the data streams were noisy. On another project, Dr. Holladay led the design optimization effort for a mass spectrometer for a space probe. This required using a specialized Particle Swarm Optimization algorithm on a high performance computing cluster to find stable operating points in a 37 dimension ion optics model. The final solution was significantly better than the best hand-tuned human solution. Dr. Holladay is also actively involved in genetic 
programming (GP) research, which included developing a new stack-based GP language, FIFTH, that operates on vectors and matrices as intrinsic data types. This advance allowed GP techniques to be applied to previously intractable problems for which data sets are best expressed as large vectors.

Dr. Holladay has been the software architect for multiple large- and small-scale projects. He was the principal architect for SwRI’s Frontier signal processing systems. These systems use hundreds of commercial off-the-shelf (COTS) computers networked together to simultaneously process large portions of RF communication spectrum. System capabilities include new energy detection, signal classification, and direction finding. Interfaces to the system are based on open standards including CORBA and XML. This work is typical of many of his projects that involve distributed computing components in both homogeneous and heterogeneous operating system environments.

Human Computer Interface design is another area in which Dr. Holladay has significant experience. As the software architect for the Frontier Remote Operating Facility (ROF), he was responsible for designing a system that collects 
high volume data from multiple Remote Collection Facility (RCF) locations, stores the data in a database, and provides operator interfaces for data analysis and visualization, as well as system management and status. The ROF operator interfaces consist of both thick client and thin client (web based) applications. He was also the PI on an IR&D project that investigated potential efficiency improvements from incorporating new HCI technologies into existing software applications.

Dr. Holladay has also worked on developing and maintaining interface standards. He served for ten years on various committees for the HART Communication Foundation, an international consortium that maintains the interoperability standards for one of the most widely used industrial instrument communication technologies. He also served for five years on the Field Calibration Technology Committee of the Instrument Society of America (ISA) where he headed the development of the Field Calibrator Interface (FCINTF) standard.

Dr. Holladay’s diverse background with real-world applications in many different domains gives him a unique cross discipline approach to problem solving. During his 40-year career, he has designed data collection, analysis, and control systems for numerous industrial processes including computer-controlled fabric dying, orange juice evaporation, municipal water treatment, metal finishing waste treatment, aluminum coating, solvent-based paint curing ovens, and industrial boilers.

Email: