11.01.2017 Views

A Technical History of the SEI

ihQTwP

ihQTwP

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Introduction<br />

In December 1984, <strong>the</strong> Department <strong>of</strong> Defense (DoD) awarded a contract to Carnegie Mellon<br />

University (CMU) to manage a federally funded research and development center (FFRDC)<br />

called <strong>the</strong> S<strong>of</strong>tware Engineering Institute (<strong>SEI</strong>). The contract award was based on a competitive<br />

request for proposals (RFP) issued to <strong>the</strong> university community in May 1984. Seven university or<br />

university/industry consortia <strong>of</strong>fered proposals, and Carnegie Mellon University was <strong>the</strong> successful<br />

proposer, in part because <strong>of</strong> its strong engineering and computer science programs and <strong>the</strong><br />

“engineering mindset” in its research efforts. The <strong>SEI</strong> is a university-based FFRDC following <strong>the</strong><br />

model <strong>of</strong> Lincoln Labs, which is part <strong>of</strong> <strong>the</strong> Massachusetts Institute <strong>of</strong> Technology. This designation<br />

distinguishes <strong>the</strong> <strong>SEI</strong> from <strong>the</strong> systems engineering FFRDCs, such as MITRE and Aerospace<br />

Corp., which are free-standing, non-pr<strong>of</strong>it corporations.<br />

Responsibility for contract management and technical oversight was assigned to <strong>the</strong> Air Force<br />

Systems Command (AFSC) at Hanscom Air Force Base in Massachusetts. While <strong>the</strong> contract remained<br />

with AFSC, responsibility for technical oversight was moved from AFSC to <strong>the</strong> Office <strong>of</strong><br />

<strong>the</strong> Under Secretary <strong>of</strong> Defense for Research and Advanced Technology ((OUSD(R&AT)) in<br />

1987, <strong>the</strong>n to <strong>the</strong> Defense Advanced Research Projects Agency (DARPA) in 1988, and <strong>the</strong>n to<br />

OUSD for Acquisition and Technology (A&T), now OUSD for Acquisition, Technology, and Logistics<br />

(AT&L), in 1998. These changes brought new perspectives to <strong>the</strong> oversight and concomitant<br />

changes to <strong>the</strong> strategic imperatives, which are discussed in subsequent paragraphs, along<br />

with <strong>the</strong> context.<br />

The DoD S<strong>of</strong>tware Environment in 1984 Motivated Formation <strong>of</strong> <strong>the</strong> <strong>SEI</strong><br />

The proposal to create <strong>the</strong> <strong>SEI</strong> originated in 1982 as part <strong>of</strong> <strong>the</strong> proposed S<strong>of</strong>tware Technology for<br />

Adaptable, Reliable Systems (STARS) program—a DoD s<strong>of</strong>tware initiative developed by a joint<br />

task force <strong>of</strong> DoD s<strong>of</strong>tware pr<strong>of</strong>essionals with support from industry [DoD 1983, Druffel 1983].<br />

At <strong>the</strong> time, DoD leaders realized that s<strong>of</strong>tware technology was becoming <strong>the</strong> enabler for flexibility<br />

and integration in mission-critical systems, but <strong>the</strong>y also recognized that s<strong>of</strong>tware was <strong>of</strong>ten<br />

<strong>the</strong> cause <strong>of</strong> system delays and failures. A task force was chartered to develop an understanding <strong>of</strong><br />

<strong>the</strong> underlying problems and propose a broad research program for DoD s<strong>of</strong>tware-reliant systems<br />

[DoD 1982].<br />

Although <strong>the</strong> DoD had traditionally been <strong>the</strong> leader in <strong>the</strong> application <strong>of</strong> computing, that trend reversed<br />

dramatically in <strong>the</strong> 1970s for a variety <strong>of</strong> reasons, including <strong>the</strong> difficulty <strong>the</strong> DoD was experiencing<br />

in applying evolving technology [Mowery 1996]. The military departments faced major<br />

challenges in developing s<strong>of</strong>tware for mission-critical systems because a significant<br />

component involved managing hardware devices (such as sensors and actuators) and control systems<br />

in real time, that is, within <strong>the</strong> cycle time <strong>of</strong> <strong>the</strong> sensor and control mechanisms, <strong>of</strong>ten milliseconds<br />

or microseconds (now nanoseconds). To meet <strong>the</strong> efficiency needs <strong>of</strong> <strong>the</strong>se real-time systems<br />

and <strong>the</strong> need for interfaces to hardware devices, s<strong>of</strong>tware for DoD systems was developed in<br />

assembly language and/or o<strong>the</strong>r low-level languages, some <strong>of</strong> which were specific to a military<br />

department or program (e.g., Jovial for <strong>the</strong> Air Force and CMS-2 for <strong>the</strong> Navy). To a large extent,<br />

programming at <strong>the</strong>se low levels precluded <strong>the</strong> use <strong>of</strong> advancing technology and tools. As a result,<br />

<strong>the</strong> DoD experienced increased development, quality assurance, and sustainment costs, and protracted<br />

schedules. In addition, low-level code was an inhibitor to <strong>the</strong> greater levels <strong>of</strong> integration<br />

needed as <strong>the</strong> DoD began attempting to field larger systems.<br />

CMU/<strong>SEI</strong>-2016-SR-027 | SOFTWARE ENGINEERING INSTITUTE | CARNEGIE MELLON UNIVERSITY 3<br />

Distribution Statement A: Approved for Public Release; Distribution is Unlimited

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!