AviationKnowledge Research Program

Institute of Cognitive Science (ICS)
University of Colorado at Boulder

Dr. Peter G. Polson, Principal Investigator
Dr. Marilyn Hughes Blackmon, co-Principal Investigator
Richard Brown, M.S., Computer Science, consultant
Karl Fennell, M.A., consultant, subject matter expert

Contact: blackmon@colorado.edu, phone 303-859-5060

Return to the home page of the PilotLSA website

Web-based LSA tools for estimating pilots' judgments of semantic similarity and familiarity:

After becoming familiar with the LSA tools, you can set up computer-to-computer queries that enable your aviation research computer program to query the LSA tools and run analyses using either the ExpertPilot or CDUskills semantic spaces. This provides just-in-time LSA similarity and familiarity values to use in your research tools.


Aviation research projects that send computer-to-computer queries over the Internet to use LSA web-based tools and ExpertPilot or CDUskills semantic spaces




Automation Interaction Design and Evaluation Methods
Sponsors: NASA Ames Research Center, Human Systems Integration Division

Dr. Bonnie E. John in collaboration with Dr. Lance Sherry at George Mason University, Drs. Peter Polson and Marilyn Hughes Blackmon at the University of Colorado, and Dr. Mike Matessa at Alion Corp, the objective of this project is to conduct foundational research to develop the underlying theoretical methods to enable the development of a new class of Computer-Aided Design (CAD) tools for the design and evaluation of Human Automation Interaction.

Specifically, the theoretical methods developed in this project shall solve two problems with existing methods and tools (1) Executable models of human performance for metrics required by the aviation industry (i) training time, (e.g. time-to-competence for a task), and (ii) operational efficiency (e.g. probability of failure-to-complete-task) (2) Automation for assessment of saliency user-interface visual cues and semantic similarity between cues and task steps. These assessments, currently performed manually, are subject to poor interrater reliability.

Previous research had demonstrated the feasibility and benefits of applying straightforward human factors methods to the design and evaluation of cockpit and UAV ground station automation (e.g. Smith & Polson, 1999, Sherry et al. 2005a, 2005b). The major problems experienced with these methods were: (i) the absence of metrics required by the aviation industry (i.e. training and operational efficiency metrics ), and (ii) poor inter-rater reliability.

The research approach proposed for this project is to: (1) augment and adapt existing executable models of human performance such as the GOMS, ACT-R and CORE in CogTool (see John et. al 2004) to generate metrics required by the aviation industry (2) extend SNIF-ACT to the aviation domain to automate the salience and semantic similarity assessment (3) demonstrate the the executable models and automation of salience and semantic similarity of real-world industry projects and perform usability analysis on the tools such as CogTool, HCIPA (Sherry et al. 2006) and ADEPT (Feary, 2006).

Alion Science and Technology The Automation Design Advisor Tool (ADAT) is being developed with NASA support to assist aviation designers, throughout all phases of design, in creating systems that ensure effective and efficient human-automation interaction on the modern flight deck. The tool can be applied to both new “blank slate” designs, and to improved versions of existing technologies. In particular, the current version of the tool evaluates proposed Flight Management System (FMS) designs in terms of their ability to support pilots in instructing and monitoring the automated system. ADAT identifies potential design deficiencies, provides specific (re)design recommendations, and offers access to summaries of empirical research on the underlying reasons for and the effectiveness of solutions to breakdowns in pilot-automation coordination.

This material is based upon work supported by the NASA. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of NASA.