Home > Uncategorized > AI Assisted Industry Training and Simulation Applications

AI Assisted Industry Training and Simulation Applications

Warbird Cyberstar

Warbird Cyberstar

Warbird Cyberstar  presents this talk about AI Assisted Industry Training within an Interactive 3D Applications Simulation. As an example, I will be demonstrating a three-dimensional Interactive 3D application simulation of a Standby Monitor Regulator Station from the Natural Gas industry. This simulation is designed for the technician to perform an operational inspection of a regulator station. Because of the versatility of the simulator the user is also able to perform other routine functions and cause and affect scenarios at ones discretion; such as closing a valve on a sensing line and stroking the primary regulators.  I have also developed an AI engine built within this simulation that essentially is an expert system that fundamentally understands what is going on and happening with the technician in real-time. Included are 3 modes of training built within this simulation containing a Virtual Tour, Training (AI) and Testing mode.

Within the training mode of this simulation (where the AI is located) it features real-time tracking of user actions and feedback utilizing the capabilities innate to its built-in expert system. In this way, the simulation provides functionality of a intelligent Virtual Instructor which is personified by a 3D avatar which follows the technician step-by-step as he or she manipulates the system. This form and function of our Artificial Intelligence is certainly applicable across a very wide-range of training situations, but what I like most about this type of training is that the intelligent instructor never grows bored and has your complete and full attention throughout the training process. The built-in AI prompts the user on what procedure or action is required at any given time. If the trainee performs an incorrect action, the AI agent notifies the trainee immediately and will “undo” that incorrect action to keep the trainee “on track” and then the intelligent instructor automatically repositions the simulation back to its previous state so the trainee can again try to perform the correct procedure or action. The instructor also provides a randomization feature which generates changes to the training procedure by randomly selecting different elements that are faulty and need repair. This keeps the training fresh and introduces a degree of uncertainty which forces the trainee to pay complete attention at all times.

Also included is an AI feature that allows the trainee to ask the instructor to provide assistance in locating gauges which are leaking and/or defective. This assistance is provided by an intelligent 3D avatar that accompanies the trainee throughout this simulation. If a trainee is unable to locate a leaking or faulty gauge and requires help, the AI instructor is intelligent enough to show that trainee exactly where these gauges are located. Not only will the avatar provide assistance at the user’s request, but also demonstrates the shortest possible traveling route to locate these gauges. I also have more adaptive programming that works along side of the regulator monitors gauges and valves that can and does effect pressuring in real time. The user can also stroke the primary regulators with a wrench and perform all the “cause and effect” scenarios at his or hers discretion, and even blow up a house down at the end of the line.

I do believe that these AI features makes this type of computer-based training more engaging, while it also simultaneously delivers greater learning retention levels for its end-users. This is a great foundational training tool and I expect that practicing on this simulator will give the user a sound foundation onto which to build upon; and to go a long way towards improving safety in the natural gas industry. The project is good example of AI assisted training within a Interactive 3D application simulation for commercial applications of Artificial Intelligence in business and industry.



Kevin Simkin’s Bio: As a committee member for the Electric Association Advisory for the Midwest Energy Association, Kevin Simkins serves as subject matter expert and liaison between virtual worlds and the energy industry. He conducts research and development into new algorithms for 3D virtual worlds as well as interactive 3D application simulations and collaboration tools. Simkins is the Ontology/Taxonomy Lead for the IEEE VW Standard Working Group. As the only individual ever to have three winning entries in the Federal Virtual Worlds Challenge, Simkins has distinguished himself as a competitive leader in the fields of virtual worlds and artificial intelligence.

Time: Saturday April 16, 2011 @ 9 AM (SLT/PDT)

SLurl: http://slurl.com/secondlife/IEEE%202/56/162/27

VyperSim: http://vypersim.com/VyperSim_Home.html

Categories: Uncategorized
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: