The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 54.0°F | Overcast

MIT To Look Into Fraud Claim Related to Missile Defense Test

By Brian Loux


MIT may formally investigate claims that scientists at the Lincoln Laboratory involved in reviewing technology crucial to the military’s missile defense system committed scientific misconduct, after a professor appointed to look into the matter surprisingly reversed his original opinion on the matter.

Professor of Aeronautics and Astronautics Edward F. Crawley investigated the charges levied by Professor Theodore Postol, a member of the Science, Technology, and Society faculty, who asked that MIT repudiate a 1999 missile defense report examining a test of TRW Inc.’s target object discrimination technology, a system similar to the Raytheon Inc. system later chosen -- over TRW’s -- for the military’s missile defense system. Two MIT scientists were members of the Phase One Engineering Team (POET) that wrote the report.

Provost Robert A. Brown last year declined to repudiate the report, calling it a “government, not MIT, document,” but said he would initiate an inquiry into the two MIT scientists involved with the report, pursuant to MIT’s policies on academic misconduct.

Crawley’s draft report said that the findings of the Lincoln Laboratory team were trustworthy. Postol protested the findings, arguing that Crawley’s report directly contradicted the findings of a 2002 report from the General Accounting Office, the nonpartisan investigative arm of Congress. Crawley then reversed his opinion in the final report issued Nov. 5, calling for a full investigation into the misconduct matter.

A press release issued by MIT Dec. 27 stated, “After reviewing the report, the Provost will determine what additional steps to take.” Brown could not be reached for comment.

Postol is on sabbatical at Stanford University and could not be reached for comment.

GAO finds flaws with report

Postol first asked President Charles M. Vest to undertake an investigation on April 26, 2001. Postol continued to request inquiries after failing to garner one from Vest, turning to the MIT corporation and members of the U.S. government.

The GAO did conduct and publicize a review on the 1999 report at the behest of Rep. Howard Berman (D-Calif.) and Sen. Charles Grassley (R-Iowa). The review, known as GAO-02-125 or the “Berman report,” released on Feb. 28, 2002, is essentially an investigation of an investigation. Scientists investigated the methods and results of the two groups and whether or not POET provided an independent and objective review.

The GAO accused the team of “not processing the raw data from [TRW’s test flight] or developing their own data by running hundreds of simulations.”

Instead, they found that the team used target signature data and reference data from TRW post-flight analysis. “The team cannot be said to have definitely proved or disproved TRW’s claim that its software [could discriminate between missiles and decoys].”

TRW whistleblower sparks case

The dispute stems from a case that began in 1996 when TRW Inc. scientist Nira Schwartz accused her company of faking test results on an anti-missile prototype system meant to differentiate between missiles and decoys. Doubts about this system would arguably harm the credibility of the overall system. Schwartz was eventually fired from her position and pursued a lawsuit against TRW.

The Justice Department and the Defense Criminal Investigative Service looked into Schwartz’s allegations. During this time, the DCIS appointed two contractors to examine the TRW findings: the Nichols’ Research Corporation and POET.

According to the GAO report, POET consisted of two scientists from Lincoln Labs, two from Lawrence Livermore National Laboratory, and one from the Aerospace Corporation. POET was used, according to the report, for fear that Nichols could not generate an objective scientific report alone.

The GAO report summarized that “[POET], which was responsible for completing an assessment of TRW’s software performance within two months using available data, found that although the software had weaknesses, it was well designed and worked properly, with only some changes needed to increase the robustness of the discrimination function.”

Based on the reports, the Justice Department said it would not help Schwartz pursue her lawsuit in March of 1999.