Executive Education:
Key to Implementing RASSP

by Jim Scharf, Sr.
and Larry Karns

Abstract

Regular readers of this Digest are familiar with the RASSP goals, the technical challenges, and how they are being met. Of equal or greater importance to the overall success of the RASSP program is the widespread adoption of the tools and methods, architectures, and infrastructure by American industry. This can only be achieved when key executives are informed of the cost/benefit issues and convinced to make the investments necessary to adopt the RASSP technology. This article describes the RASSP Education and Facilitation (E&F) team efforts to document and present the business benefits of embracing RASSP achievements.

1. Introduction

The RASSP program is now in its fourth year and has a significant list of achievements to its credit. These results can be accessed on the RASSP website at http://rassp.scra.org - one of the three highest rated websites in DSP.

RASSP is a program which was intended to benefit a wide segment of the American Electronics Industry. To ensure that the advancements developed under RASSP find their way into industry and academia, DARPA has funded a separate major task under RASSP for the purpose of Education and Facilitation to focus on:

• University Education
• Executive Education
• Single source of information.

The Information activity of the RASSP E&F program has the goal of disseminating information about the RASSP program and related activities. Mechanisms such as this Digest and the RASSP web site are used to accomplish this. The RASSP web site provides access to RASSP program information and links to many simulation models for digital circuits.

The Executive Education task has the goal of creating the technology pull or demand for RASSP methodologies and engineers trained in them. The targets for this effort are threefold. First, it is industry which creates jobs and builds products, and the adoption of the RASSP methodologies and approaches will increase the demand for the supply of skilled engineers. Second, purchasers of these needed services and systems, namely the government project and program offices, will help create demand if they request systems designed in the RASSP manner. Finally, demand can be created at the specific defense management colleges by training government program managers in the advantages of acquiring systems based on RASSP. This article focuses on the Executive Education activity of the RASSP E&F Program.

2. The Need for Executive Education

It is necessary to educate key executives and decision makers about the benefits that they can derive from RASSP adoption. To successfully educate any audience, the needs of the customer must be addressed in order to get and hold his/her attention and maintain interest while conveying information.

One goal of RASSP E&F is to communicate with executives who manage people that design digital signal processors and systems or who manage the procurement of those systems. They need to be informed of the benefits that they and their companies or agencies can derive from RASSP. In order to have the maximum impact on executives, a familiar venue with which they can relate needs to be utilized.

3. The Executive Education Approach

Educating the DSP community is necessary in order to create the demand for RASSP. Presentations or seminars are given to key individuals in the target organization. The RASSP E&F team works hand in hand with the RASSP prime contractors to ensure maximum success of these seminars. The components of the Executive Seminar are:

• Management Presentation including the business case for adopting the RASSP methodologies, architectures and tools (infrastructure),
• Top level technical presentations, and
• Technical demonstration(s).

The demographics and desires of the target audience define the structure of the executive seminar (i.e., agenda for the day). The day can be the Management Presentation alone. This will be done by a member of the RASSP E&F team and can be accomplished with or without the attendance of one or more individuals from the RASSP prime contractors. Our experience has shown tremendous advantages when a member of the prime contractor's team is present to answer any of the more technical questions. This presentation, with questions, lasts about two hours and is aimed at the high level decision makers of an organization.

The second option is to add technical presentations, presented by a member of the RASSP prime contractors, to the management, or business case, presentation. This will add another couple of hours to the Management Presentation, so this Executive Seminar would last a half day. While the first part of the presentation is aimed at the high level decision makers in the organization, the target audience for the technical presentation is the technical manager and lead engineer.

A company, or government program office, can elect to have a day of RASSP by also adding technical demonstrations for the afternoon. The audience for these demonstrations would be similar to that for the technical presentations. If this option is chosen, there needs to be a great deal of discussion among the target audience, the RASSP prime contractors and the RASSP E&F team representative. There are many tools and parts of the methodology that can be demonstrated during this half day. An entire demonstration of everything involved in the RASSP program could take a week or more, so it is imperative that the major interest of the target audience be understood to maximize the benefit of the time spent during these demonstrations.

4. The Management Presentation

The purpose of the Management Presentation is to present the business case for the adoption of the RASSP methodologies, architectures and infrastructure. The presentation is organized into the following sections:

• Problem statement
• Efforts already being done outside of RASSP
• The need for RASSP
• The RASSP technical approach
• Results to date
• Business case analysis
• Next Steps.

4.1 Scheduling a Seminar

The process of scheduling and setting up an executive education seminar begins with suggestions of candidate organizations and initial contacts and telephone inquiries to validate the choice of organization, obtaining Government approval, and locating the proper facilitator at the candidate site. Any organization can become a candidate for an executive seminar by contacting the RASSP E&F program representative, Anthony Gadient, (803) 760-3376. Candidate organizations need to be involved in the development of digital signal processing, either for military or commercial markets, or they should be government representatives involved in the procurement of DSP systems. The facilitator at the site will be responsible for arranging a meeting with high level executives at his organization.

4.2 Prior to the Seminar

Prior to formalizing the agenda and schedule, the prospect is encouraged to visit the RASSP www site (http://rassp.scra.org) and to view a 15 minute video tape overview of the RASSP program produced by one of the prime contractors.

If only the single executive briefing can be achieved in the first meeting, all efforts are made to encourage follow up at a later date at the customer’s site or at the prime vendor’s site for the full technical lecture and demonstration(s).

4.3 Seminars Delivered to Date

To date the Management Presentation has been delivered at eight major organizations, including Texas Instruments, GEC Marconi, and Rockwell/Collins. Additionally, the Management Presentation has been delivered at government installations, such as Tinker Air Logistics Center.

4.4 Feedback

Subsequent to the delivery of an Executive Seminar, the attendees are asked to fill out a simple one page, two sided evaluation form regarding the achievement of objectives and to solicit ideas on how the experience could be improved for the attendee. Responses received to date have been positive, and a number of good suggestions have been offered which will be incorporated into future presentations.

One area that is being improved is the business case for RASSP. Executives are attentive to messages transmitted in terms of profits and losses, assets and depreciation, expenditures and savings and overall return on investment (ROI). While experiential data has been used to support the business case, it is very hard to come by good, independently verified business case data. The RASSP E&F Executive Education team is working diligently in seeking out good data and any feedback or suggestions would be welcomed.

The RASSP program itself is producing a significant volume of verified cost data. One of the key elements of the RASSP program is the development and measurement of specific performance improvements through benchmarks. These benchmark projects are defined and monitored by MIT Lincoln Lab, an entity independent of any of the RASSP prime contractors or tool vendors.

5. Back to Business

The current presentation already contains convincing evidence of the business reasons to adopt the benefits derived from RASSP. A basic business example from the current presentation is shown in Figure 1. This graph shows that in a commercial setting, a delay in bringing a product to market within a proper “market window” of time can cost a company in lost revenues and consequently lost profits. The case is made that RASSP accelerates the development process significantly, thus averting the potential loss.

DARPA’s thrust with RASSP can have even more impact for military prime contractors. As shown in Figure 2, the typical scenario for systems development is centered around achieving a significant level of performance beyond current capabilities. This typically involves trying to push technology too far in the early years of a program with a point solution in order to achieve (possibly overly ambitious) goals, only to see system performance ultimately fall behind the commercial development curve in later years. The cost of the single point push at the inception of the project is high and artificially inflates systems cost, while the loss of capacity in the later years drives the early need for a replacement system - again at higher cost.

By contrast, RASSP’s Model Year Architecture approach alleviates the problems on both ends of the time line, saving a great deal of cost. With the RASSP Model Year Architecture approach (displayed in Figure 3), a scalable, flexible approach is taken in the initial design phase, and all of the information is captured in a database. As time progresses, lessons are learned from the initial design, and the technology curve continues to advance. With a high degree of reuse guaranteed, it is far more cost effective to roll out improved designs in succeeding model years than the traditional “push hard, then age” scenario.

Figure 4 shows how the costs of a project are committed early, long before a physical prototype can be produced. Once these costs are committed, it becomes too late to run “what if” studies or to explore alternative architectures and methods which might have improved the system performance or reduced costs or both.

Through the process of Virtual Prototyping, hierarchical models of trial systems can be built, explored and evaluated in a software virtual prototype. Actual data showing the benefits which can be achieved by the use of Virtual Prototyping, when presented as in Figure 5, makes a powerful case that is easily understood by engineer and executive alike. Direct enumeration of the actual average times required on a RASSP based system, which has been reduced from months to days, conveys a strong and unmistakable message. Graphics, such as these, coupled with compelling stories of actual use of RASSP techniques, help to make a forceful case for the executives who make the business decisions.

6. Continuing Improvement

Work is continuing with the RASSP prime vendors and MIT Lincoln Lab to develop additional business cases and amplify and focus the messages from the cases already in use. Collaboration is being actively pursued with other companies who have case history data for work that they have done using methods that are subsets of RASSP. Additional information is being sought from companies who wish to adopt some or all of RASSP’s methodology, infrastructure and architecture and who are willing to participate in “before and after” case studies for future publication.

One significant piece of additional work that has recently begun is the development of a formal business case example for the adoption of RASSP by a DSP systems vendor. The initial case chosen for study was one developed by Lockheed Martin ATL under their contract to study cost models. They have produced a detailed comparison of the development of a new AWACS-like system with significant DSP requirements both under the RASSP methodology and in the conventional manner. This report provided the basic facts required to build a business case.

The model for building the actual case came primarily from the Department of the Army, "Economic Analysis Manual", U.S. Army Cost and Economic Analysis Center, July 1995. The results of the analysis showed that it was possible to utilize the cost savings from switching to the RASSP approach on the demonstration-validation phase alone to pay for all required tooling and training required by the adoption of RASSP and still produce an acceptable return on the initial investment (Table 1). In this case, the traditional methods would cost $12.9 Million for the Demonstration/Validation phase. Adopting RASSP techniques would require an initial investment of $2 Million, but would return 103% on the investment in two years.

Updates to the Executive Seminar presentation will be made as new information evolves. Likewise, the RASSP E&F Program will continue to make improvements to better achieve the goal of promulgating and proliferating the benefits of the RASSP program to American industry and government.