ucfcollege of sciencespsychology | institute for simulation and trainingtpl

welcome| research| publications| people| downloads| links| location| contact


Current Research Efforts

Human-Robot Interaction Research

TPL is currently a participating member of the Robotics Collaborative Technology Alliance (RCTA) investigating humanórobot interaction in future humanórobot teams. The RCTA is an interdisciplinary consortiumófunded by the U.S. Armyócomprised of engineers, computer scientists, cognitive scientists, and human factors psychologists from public, private, and academic institutions. Together, these researchers are working on issues related to the design and implementation of future robotic teammates intended for military purposes.

TPLís research within the RCTA primarily examines two concepts: situation awareness (SA) and shared mental models (SMMs). Below you will find the three RCTA research projects that we are currently undertaking to examine these two concepts:

Model and Improve Mission-Level Situation Awareness

The objective of this research is to support peer-to-peer tactical teaming by informing management of humanórobot communication at the mission planning, cognitive level. Specifically, the goal of this research is to understand how capabilities of a robot can support Soldier situation awareness regarding the robot. Our current work under this project involves investigating how factors related to robot-to-human information exchanges ( i.e., information sent from a robot to a human team member), impact Soldier situation awareness. Factors under investigation include frequency, medium (e.g., visual, auditory), and information presentation.

Determinants of Shared Mental Models for SoldieróRobot Teams

The objective of this research is to support the development of peer-to-peer tactical teaming and fast, adaptive learning by expanding shared mental model understanding between human operators and robotic systems. Specifically, we are interested in assessing the mental models that humans hold of robots. Our current work under this project involves assessing human expectations of a robot teammate as it autonomously performs commands (e.g., navigate, search, observe) under different constraints (e.g., quickly, covertly) while considering features present within the environment(e.g., bystanders, enemies, difficult terrain).

Instantiating HumanóCompatible Computational Mental Models of Shared Mental Models in SoldieróRobot Teams

The objective of this research is to support fast, adaptive learning and peer-to-peer tactical teaming by enabling robots to support tactical decision-making in dynamic op-tempo environments. Specifically, we support these goals by providing relevant task decompositions (e.g., "Screen the back door") that are used to generate decision models that produce tactical behaviors (e.g., Navigate, Search, Observe), which support Soldieró robot teaming across a variety of mission types and op-tempo environments. These decision models serve as computational mental models that can be instantiated within robots.

Past Research Efforts

Past FAA Research

Identifying CRM Approaches for Enhancing Flightcrew Performance

TPL assisted the Federal Aviation Administration (FAA) by performing research to identify, and assess the current state-of-the-art in Crew Resource Management (CRM) approaches, tools, and techniques.
The approach taken by our lab consisted of two phases of research:

1) Reviewing the existing CRM literature, including the concept of Single Pilot Resource Management (SRM), to identify and assess state-of-the-art CRM approaches, and provide data that could be used by the FAA to update AC 120-51E as well as advance the state of CRM

2) Developing and conducting an empirical study to assess the efficacy of selected CRM approaches with respect to flightcrew performance (e.g., the improvement of problem-solving, decision-making, situational awareness, and team communication skills) in an area of application.

Improving Pilot Training for the Operation of Automated Aircraft

This research effort was developed for the purpose of furthering our understanding of mental models and how that understanding can better the processes needed to train those who operate systems that require human monitoring and intervention. We were specifically investigating what type of knowledge should be contained within a pilot's mental model of automated flight deck systems (i.e., knowledge about the system's components, the system's operation procedures, and knowledge of the relationship between these two).

Threat and Error Management

We examined what types of knowledge affect an operator's ability to react to abnormal or threatening events while operating a complex system. The goal of the research was to determine whether a greater focus in training on the conceptual understanding of threatening events would lead to an improvement in flight crew performance, especially in threat management.

Past TSL Research

Our research lab was involved in a multiyear project sponsored by the Transportation Security Laboratory (TSL) under the umbrella of the Department of Homeland Security (DHS).

After four years of successful research looking at the underlining dynamics behind a complex visual search task such as the X-ray screening of bags, our lab developed a unique training approach aimed at further improving the detection rate of threatening items such as guns, knives, and improvised explosive devices (IEDs)

We had two main research foci:

1) Developing a field experiment in which our latest training paradigm was tested on airport X-ray screeners around the country in order to determine if the findings in our laboratory could extend to such population. This applied research was the first step to validate our training protocols.

2) Furthering our laboratory exploratory research to precisely look at the impact factors such as differential clutter overlap could have on threat items, the respective repercussions on detection rates, and how to mitigate those effects.

Past ARL Research

Our lab conducted research that was funded by the Army Research Laboratory. This initiative focused on the operation of unmanned vehicles and considered the influence of factors including teams, spatial ability, mental models, environment, and skill acquisition on performance.

Performance of Distributed Teams Controlling Mixed Assets

One of the studies that was being conducted under this project examined the performance of distributed teammates controlling an unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV). The UAV operator controlled the asset from a mounted office environment, while the UGV operator controlled the asset from a dismounted foxhole environment. In addition to basic communication, this study considered the effects of sharing a teammate's vehicle view and sharing control of a teammate's vehicle on coordination, workload, and performance.