FrogOJT Systems

O/ 510-843-6227 FAX 510 843 6280 H/ 510-620-0512/Pgr 510-425-2933

A New Approach to On the Job Technical Training for Critical Repetitive Physical Tasks, employing digitized video for structured OJT.


A detailed look at Digitized Video Structured OJT

We made a proposal to the Federal Government in 1995 to develop our approach to structured OJT. They said it wasn't "novel". So we went ahead and did it ourselves, and then tested it on young hockey players at the Oakland hockey rink. You will see how short these things actually are [<5 sec.]). We worked out a lot of the technical aspects referred to in the text below. I satisfied myself that skill acquisition does in fact take place, and at an astonishing pace. 11/10/97 RK


EXCERPTS from SBIR Proposal


This Small Business Innovation Research Phase I project would develop the concept of an innovative method for structured on-the-job training--digitized video interface--and would analyze the cost-effectiveness of this training system in pilot projects. The development of this training method addresses an important business problem, namely that a large proportion of on-the-job training currently being conducted by US companies is less effective than it might be. The result is lower productivity, higher costs, and lower quality than would be possible using a better system of structured on-the-job training. The research program will have two parts. First, experiments will be conducted to identify the essence of powerful video presentation for the imparting of skills in order to develop and produce the training interface. Second, pilot training projects will be conducted and data collected to analyze how to use this interface effectively in structured OJT. The goal is to develop principles to guide the production and then implementation of the interface in a cost effective manner in the workplace. Phase I will develop the concept of the interface and test its effectiveness as a training tool.





The commercial application of the research is production of a computerized training system for businesses that use on-the-job training for the development of demonstrable manipulative skills. The uses of digitized video training material in commerce extend to all applicable physical skills. The number and distribution of commercially valuable physical skills is vast.




This proposal develops the concept of an innovative method for structured on-the-job training, or OJT. Previous research has shown that structured OJT is the most cost-effective form of training in certain circumstances (Jacobs, Jones, and Neil, 1992; Rothwell and Kazanas, 1994). However, US firms usually engage in formal training, mostly in scheduled classroom-type settings; and in non-structured, or haphazard, OJT, mostly in production settings, during work hours. The primary training resource, if one is available, is detailed training guides. Research has shown that companies put considerable resources into training and suggests that this training could be made more productive (Brown and Reich, 1995; Bishop, 1994; BLS, 1994; Rothwell and Kazanis, 1994; Carnevale and Gainer, 1989).


What is proposed here is a new form of structured OJT at the production site, using recently available techniques employing digitized video, which provides a worker/trainer/virtual-task interface, which can be accessed and controlled by the worker. Phase I of the project will undertake research on the optimal ways to digitize specific demonstrable manipulations for structured OJT. In addition, pilot projects in selected work sites using the training tools will be used to measure their success in skill acquisition in terms of the time required, the costs of training, and the performance achieved using this method of structured OJT.







The typical delivery system (interface) for production-sited OJT has been worker/trainer/task, followed by worker/task. That is, first the trainer teaches the worker how to perform the task, by demonstration and explanation. Then the worker acquires skill and reliability by performing the task. Both classical and contemporary research in perception and learning have indicated that humans (and other organisms, even the octopus) can acquire knowledge by mere observation. The stumbling block, and a major determinant of cost, in OJT is that the number of observations is limited by the number of repetitions the trainer has time, patience, or inclination to perform. Now, using digitized video, we can have unlimited repetitions by skilled hands of demonstrations of critical tasks (A critical task is herein defined as one that costs time and money, and upon whose successful completion depends the savings of much more time and money . Failure in a critical task precipitates a cascade of losses. )


This type of training system should be especially useful in teaching specific, manipulated, repeated work tasks that are critical to the work process. The interface should be especially successful with noncollege goers who learn their primary employment skills at the workplace. The proposed computer interface should provide employers with a valuable system to train and document the skills for this group of workers as part of the Apprenticeship 2000 Review and the National Job Analysis Study (ETA, 1989; ACT, 1994).




What we propose is a worker/trainer/virtual task interface: A trainer's movements are videotaped and digitized and then compressed into a Quicktime® movie. The trainee is given access to a Macintosh computer which has the movie loaded onto its hard drive. The trainee is given the freedom to watch the movie -- the movie of just the figure-eight style movements to finish off fiber optic cable, for example&endash;&endash; a critical task1 in the telecommunications industry -- as many times as desired, as many times as are necessary to imprint that particular skill. The trainee then tries to perform the task in the field. The trainee may return to that digitized video as many times as desired.




This research program has two objectives. First, we seek to discover the essence of powerful video presentation for the imparting of skills in order to develop and produce the training interface. Second, we seek to analyze how to effectively use this interface in structured OJT so that the interface is produced and then implemented in a cost effective manner in the workplace. Phase I will develop the concept of the interface and test its effectiveness as a training tool. Phase II will allow the refinement of the concept and the development of low-cost methods of production. Phase III will focus on the production and delivery of the interface product to businesses for use in their structured OJT.


While every technician in a craft acquires their own style for accomplishing that task, a trainee needs to emulate at least one competent rendition in order to become minimally proficient -- to be able to attempt the task on his/her own. Currently, a trainee gets to watch a trainer perform this task a number of times. This is the traditional worker/trainer/task interface. However, the number of repetitions observed is dependent on the trainer's patience, and subject to the pressures of getting the work out. Traditionally ,the trainee then goes to the worker/task interface, learning by doing, turning out product under the limited supervision of the trainer.


The Feasibility of the Interface: Establishing the Principles for Effective Presentation


Developing the worker/trainer/virtual-task interface will require onsite experimentation for effective acquisition of the essential visual and aural inputs. Which part of the sound spectrum is the critical sound to record and train by? Which point of view should the camera take to display all the proper body mechanics? What is the most effective viewpoint for the trainee to "look over the shoulder" of the virtual trainer? Should the digitized and postproduced video be one shot, with no cuts or closeups? Is a larger screen image with lesser resolution more effective or less effective than a smaller screen image with higher resolution, i.e., less pixellated.


If there are specific principles for the production of effective digitized video presentations, we shall discover them and codify them from the repeated acquisition of such materials across varying fields of work, beginning with generally accepted US production techniques and the generally high production values of US broadcast video, and evolving techniques for digitization and post production of Quicktime® movies.


Why not just use video? The exhibition medium to be developed here is different qualitatively from simple video training materials. These digitized videos can be played on the Macintosh, "rewound", and played again, instantly. To do this on a VCR is too cumbersome to be practical. Digitized training videos are short. They are concise and unadorned. They are nonlinear. A trainee can inspect them a few frames at a time, or a single frame at a time, or at 30 frames per second for the lifelike appearance essential to learning a task.


Effectiveness of the Interface in Structured OJT:


Protocols will be drawn up for the systematic testing of varying video presentations for a sample of tasks across a number of fields. The results will be noted in terms of worker skill acquisition, gains in productivity (or cost reduction) through the structured OJT, the time it takes for a worker (new and experienced) to internalize the recommended demonstrated method, the amount of supervisor time or other workers' time used in the training process, the cost of the training system in both direct and time costs. To the extent possible, we will compare these cost and performance outcomes with the previously used training method.


Implicit in this is the cooperation of the clients in developing digitized video structured OJT.




The use of digitized video training material in commerce probably do not extend beyond commercially applicable physical skills. Tasks requiring an intellectual or clerical approach -- training in computer software applications in an office, for example -- probably will not be amenable to such instruction. Nonetheless, the number and distribution of commercially valuable physical skills is vast. Each one of these skills is calling out for its digitized archetype, to be shot and disseminated to all of the aspiring practitioners of that particular art --worldwide. There is no language barrier to watching a Macintosh.


Given that firms typically spend up to 2% of their payroll on various types of training, the commercial potential is large.




The principal investigator is Richard Katz, 47, President of Frog's Rentacomputer of Berkeley. Mr. Katz received a Bachelor of Arts at the University of Pennsylvania majoring in biochemistry, and went on to do graduate work in biochemistry at the University of California at Berkeley.


Mr. Katz has worked in a number of fields, including biochemistry, construction, trucking, and computer rentals. He has experience in writing and producing plays. His experience in the business and entertainment worlds has led him to the current project of analyzing the presentation and production of digitized video for structured on the job training. Through his theater work, and through contract work with Hollywood production companies for computer services, he has formulated the proposed interface and wants to prototype it and test it under as rigorously controlled conditions as feasible, given that it entails working with people's various learning styles and abilities.


James E. Helms is a Macintosh Specialist, skilled in all phases of Macintosh applications software deployment and with all the various configurations of Macintosh hardware and third-party peripherals. He is highly qualified to venture forth into the proposed research, where success will be dependent on the development of an innovative system that requires integration of hardware, application software, and content.


Dr. Clair Brown, Professor of Economics at the University of California, Berkeley, will serve as an advisor to the research project. She has extensive experience studying training systems in US and Japanese companies. Dr. Brown has been the director of a research team of academics from the US and Japan who have been engaged in field work at dozens of US and Japanese companies over the past seven years. A draft manuscript of their research findings is titled Employment and Wage Structures in US and Japan. Dr. Brown is also Director of the National Center for the Workplace, which is a consortium of five universities that conducts research on high-performance workplaces and is funded through the US Department of Labor. Dr. Brown will develop the fieldwork instruments for collecting data and analyzing the cost-effectiveness of the pilot training projects.



For the portion of the proposed research that involves videography, and all of the attendant adjuncts of video production, freelance subcontractors will be hired on a daily or "til completion" basis, as is customary with video and film production. These subcontractors will include: Gaffer and assistant; Transportation; and Camera Operator and assistant. These are routine services from readily available commercial sources. None of this turned out to be necessary, as long as the facilities people made sure all the lights were working.




The equipment required, and currently owned, which will be utilized for this project on an exclusive basis, includes:

For Postproduction:

Macintosh 8100/80, 80 Megabyte RAMwith two 4gigabyte hard drives;

Radius VideoVision Studio;

20" RGB monitor, 17" monitor, and NTSC monitor;

For Production:

Hi8/ SVHS video camera;


Two studio halogen lamps and batteries, overhead tracks, bluewall;

For Playback:

Macintosh with 24 bit video card;


The above is all rental equipment that will be taken off active rental and assigned to the proposed research account.


(Want to download a typical FrogOJT structured-OJT-style video?)


ACT 1994. Performing a National Job Analysis Study: Overview of Methodology and Procedures. Iowa City: American College Testing.

Bishop, John 1994. "Employee Training in the United States: a Review of the Literature."

Background paper for OECD. Paris: OECD.

Clair Brown and Michael Reich 1995. "Employee Voice in Training and Career Development," Paper presented at the Industrial Relations Research Association annual meeting, Washington, DC, Jan. 1995.

BLS 1994. "BLS Reports on Employer-Provided Formal Training," Press release, Sept. 23, 1994, Washington, DC: US Bureau of Labor Statistics.

Anthony Carnevale and Leila Gainer 1989. The Learning Enterprise. Washington, DC: US Department of Labor.

ETA 1989. Work-Based Learning: Training America's Workers. Washington, DC: US Employment and Training Admin