Successful construction projects live or die by the quality of the schedule, yet some 20 years after scheduling software became the dominant tool for tracking a project's critical path, experts say it's still not to be trusted.

Andy Ness, a construction attorney for law firm Jones Day, often talks to industry groups about the trouble he finds at the intersection of law and project scheduling software. In his opinion, legal disputes that rely on forensic analysis of a schedule are too often thwarted by "smoke and mirrors" software features, which can undermine the Daubert standard that governs expert testimony.

Ness' issues are threefold: Many scheduling experts who are called on to testify are not qualified; scheduling software allows users to manipulate a desired outcome; and, perhaps most important to attorneys, critical path method (CPM) schedule analysis can be incapable of yielding useful results about job delays.

"This may not be fair [to the many qualified scheduling experts in the industry]," he concedes, but "the problem is, there is no clear standard, no degree or requirement to be a CPM expert." If a case turns on a technical area of the schedule that judges have never encountered, "some are interested to find out, but a lot of them aren't," he adds. This is one reason he says many project-delay cases are settled instead of going to trial.

Ness says scheduling features such as user-assigned constraints can be abused and obfuscate a project's critical path. In addition, "progress override" and "retained logic" features in the software can help users make a project look like it is on time when it's not.

Attorney Stu Ockman of Ockman & Borden Associates points to one example: a major offshore petrochemical project that had 528 early-start constraints and over 400 late-finish constraints within a network of about 2,900 activities. The multiple constraints made finding the critical path for the project's start and end dates impossible, not to mention the nearly 83 workdays of negative float they yielded. Lawsuits followed the project.

But Richard "Dick" Faris—a co-founder of Primavera, now owned by Oracle and a dominant scheduling tool in use—has heard these arguments before. "This story is as old as scheduling itself," he says of the complaints (see ENR's 5/26/2003 cover story). "[Primavera] has evolved to meet the needs of customers, and the lawyers aren't happy that it doesn't make it any easier for them to prove certain things," he counters. "We're making the software for people who are building scheduling models, not for people who do forensic analysis."

After all, he adds, anything can be manipulated. Multiple calendar features, retained logic and overrides are necessary in real-life schedules. Owners need to be clear with contractors how they want to use the scheduling software, he says. "If they want them to use retained logic, they need to say so in the beginning."

In addition, Faris says, "It's up to the scheduler to decide which [scheduling features] to use, and it's up to owners and general contractors to dictate to their contractors what logic they're supposed to use."