But in the 20 years since the article, it seems that software developers have failed to deliver a rigorous protocol to mathematically calculate “truthful” schedules or at least more nuanced versions of them. Their efforts, understandably, focused on expanding market share by trying to sell software to big program customers who would require all firms on a project to use their product.

An immediate impact of the ENR article was to transform two separate (and often competing) industry groups, as both sought to address issues raised. The Project Management Institute, founded by early adopters of CPM planning and scheduling, had by 2003 become more concerned with cost control and estimating and started a College of Planning & Scheduling.

But over time, the institute and its recently created college inevitably clashed over control and policy when it came to ethical scheduling practices. Institute leaders demanded that those running the college support the institute’s definitions of terms and practices.

After the college’s leaders refused to comply with the institute’s demands, it eventually shuttered the college. It instead used a community of practice approach I believe tends to rely on the institute’s copyrighted project management body of knowledge.

In roughly the same period, the Association for Advancement of Cost Engineering International shifted somewhat from its growing focus on cost and estimating back toward planning and scheduling.

It then became clear to me that training, certifications and recertifications of the two groups’ endorsed practices would be used to pay for their activities, as is typical of most technical or membership associations. Software makers and third party trainers also saw opportunities for financial gain, which is their purpose.

The association did set up a contract and dispute resolution subcommittee to study the validity and stability of various software programs used for forensic analyses. What it found was troubling. In some cases, the same data fed to two devices created differing results that software developers could not explain. Even worse, some data entered into the same devices differed depending on which day it was run. Many papers were written, peer reviewed and published.

A subcommittee member drafted a “Letter to the Court” that would, in the case of a lawsuit, explain how CPM software used on the same project for delay analysis could produce differing results. But the letter was tabled in 2019 and never approved. Several subcommittee members raised concerns that many of their association peers had too much invested in one software program, and that the letter could be financially disastrous.

A common thread of these developments is that the very definition and algorithms used in CPM training and software had been geared not to produce honest appraisals of who or what caused a project delay. Instead, they were geared to help parties to a delay disputes win. That different software still can be manipulated to “calculate” the responsible party—even when liquidated damages may exceed $50,000 per day—just is not right. If used forensically, this could be seen as misleading the court.

This brings me to the Construction CPM Conference I started 10 years ago. It was meant to fill the void left by the groups and to serve as a place where free debate can occur. It has become a gathering place, I’m happy to say, for the top practitioners of planning and scheduling, and users of all the software. They can then discuss the how and why of practice and training when using the powerful scheduling tool.

The issues aired in the ENR article 20 years ago remain unresolved, so I hope to see you at the upcoming CPM conference in San Antonio, Texas, which is set for Jan. 14-17. Together we can continue the unfinished work we must do to perfect the practice of scheduling and planning.