Jeremy Ludwig

An AI Modeling Tool for Designers and Developers

We describe an AI modeling tool meant to be used by both designers and developers. The method for authoring is visual and meant to convey decision logic in a more intuitive manner while retaining expressiveness. This data-driven approach features an AI runtime engine which incorporates several augmentations which make it suitable for use across a wide array of deployed systems.

PC Rapid Modification Tool for Aircraft Experimentation & Training for the MH-60S/MH-60R Helicopters

The US Navys PMA-205 in conjunction with Stottler Henke has developed and deployed a tool that can be used for training and AOP experimentation for the US Navys new MH-60S and MH-60 R helicopters. The tool, called the Operator Machine Interface Assistant (OMIA), is primarily an expandable, easily modifiable low-cost PC-hosted desktop crew trainer. OMIA is currently in use for training at HSC-2, HSC-3 and HSM-41. However, since both helicopters have been constantly evolving through the development process of OMIA and continue to evolve now and in the future, OMIA had to utilize flexible, low-cost, rapid development methods.

Rapid Simulation Construction

In the course of building four low-cost and fairly distinct desktop training simulations for the Air Forces Air University and a simulated control system for NASAs International Space System, we have worked to develop in-house standard practices and a toolset for rapid construction of training simulations. We present here this process and toolset and discuss its key strengths and shortcomings in the context of the simulations we have built using it. We focus in particular on two aspects of development that we found to be pivotal: scoping the level of fidelity for the simulation logic, and designing and constructing cost-effective simulation user interfaces that achieve instructional goals. These two aspects will be described in the context of the SimVentive and SimBionic toolsets, both of which are freely available for use by NASA or any other government agency.

ITADS: A Real-World Intelligent Tutor to Train Troubleshooting Skills

Real-world intelligent tutoring systems are important ambassadors for promoting wide adoption of the technology. Questions about affordability, quality control, operational readiness, training effectiveness, and user acceptance are significant in this context. This paper describes ITADS, an intelligent tutor developed to provide a problem-based, experiential learning tool to complement schoolhouse training. The goal was to train US Navy Information Systems Technology support staff in troubleshooting skills through the use of realistic simulations and automated assessment and feedback. This paper describes the tutoring system and a preliminary validation study of its training effectiveness. The results demonstrate that the system is effective in improving troubleshooting knowledge and skills. The ITADS system was successfully developed in twenty-six months from requirements to validation, following strict systems engineering procedures. The results of the training effectiveness study indicate that the ITS also leads to significantly improved performance among Navy IT recruits in troubleshooting tasks.

Static software metrics for reliability and maintainability

This paper identifies a small, essential set of static software code metrics linked to the software product quality characteristics of reliability and maintainability and to the most commonly identified sources of technical debt. An open-source plug-in is created for the Understand code analysis tool that calculates and visualizes these metrics. The plug-in was developed as a first step in an ongoing project aimed at applying case-based reasoning to the issue of software product quality.

Intelligent pairing assistant for air operation centers

Within an Air Operations Center (AOC), planners make crucial decisions to create the air plan for any given day. They are expected to complete the plan in part by pairing targeting or collection tasks with the available platforms. Any assistance these planners can acquire to help create the plan in a timely manner would make the entire process more efficient and effective. This paper describes the Intelligent Pairing Assistant (IPA) prototype, which would provide pairing recommendations at specific decision points in the planning process. IPA is designed as a plug-in for software systems already in use within AOCs. The primary contribution described in this paper is the application of existing research in intelligent user interfaces to a novel domain.

MH60S/R helicopter multi-platform & Web-based crew trainer with FLIR

The US Navys PMA-205, in conjunction with the training and simulation industry, has developed and deployed OMIA: a flexible, multi-platform, Web-based crew trainer for the Navys new MH-60S and MH-60R helicopters. 12OMIA is currently in use by HSC-2, HSC-3 and HSM-41 and is available to all crewmembers throughout the Navy. To maximize access to the trainer and to make deployment easier, OMIA is written in Java, and deployed both as a portable application and as a web application. A portable application does not require any special user rights to install; a web application is run over the internet from a browser. OMIA includes a simulation of the MH-60S/R Common Cockpit, including a FLIR capability that includes using an actual hardware FLIR hand-control unit when attached through USB. OMIA has been designed and implemented to be flexible to changing Navy needs, a design aspect which proved itself again in 2009 when OMIAs FLIR was converted for use in the fixed wing EP-3E aircraft. OMIA illustrates solutions to three critical issues in developing low-cost training software for aircraft: quickly responding to the ever-changing demands, re-use of interface components across multiple aircraft, and providing training that can be accessed where and when it is needed.

Developing an International Space Station curriculum for the Bootstrapped Learning program

DARPAs Bootstrapped Learning (BL) program is aimed at advancing the state of the art in instructable computing. Two objectives of this program are developing a general electronic student that makes use of machine learning algorithms to learn from the kind of focused instruction typically provided by a human teacher and creating a repository of automated curricula that can be taught to the student. This paper focuses on the second objective, describing a curriculum developed for the BL program to both instruct and test the student that places the electronic student (eStudent) in the role of an International Space Station (ISS) flight controller. The eStudent is taught how to detect and diagnose single-fault problems within the thermal control system of the ISS. During each lesson, the eStudent interacts with an ISS simulator to review alerts and access telemetry values. To obtain greater visibility into its diagnostic reasoning, the eStudent is trained to create an external representation of its reasoning about the current problem - a diagnostic rationale. This includes describing potential problems, hypothesizing possible events and states, positing possible causal explanations as rationale assertions, seeking evidence for or against these assertions, projecting possible risks, and using possible risks to focus attention when developing a rationale. In addition to describing the curriculum developed as part of the first year of the BL program, we also describe some of the future directions we will investigate as part of the second year. 1,2