Anticlue

Lessons Learned - learn the lessons

I've been following the tale of the VA in St. Petersburg, FL with great interest. It is a tale of woe and hardship and the implementation of HealTheVet at the Bay Pines Medical Center in St. Petersburg, FL.

From an outside view of one who only really reads the newspaper articles, it appears that last years fiasco surrounded staging the implementation of the CoreFLS, the financial management, and then following up with HealtheVet a classic two phase implementation.

Well CoreFLS, was a failure of a $372 million proportions, and is still under investigation. Carnegie Mellon was asked to review the upcoming implementation of the HealtheVet.

Their findings:
First, it will take $3.5 billion to modernize the VA with the HealtheVet System. (Problem only 311 million was budgeted this year) By the way, these are not in order of importance.

Second, even if they spent the money and create the system it will not do what is necessary. The project lacks a leadership vision, has poor communication based on subjective information. In addition the to be state has been scaled way back and critical functionality is eliminated to stay on time. There is no defined current process and there is no defined future process. Change management and training is not fully grasped by the leadership of the project. Sometimes the basics are necessary.

Now in reading this, what is really interesting is that the project is doomed. It doesn't matter if the code has Procedural, OO, used a framework didn't use a framework, was SOA or wasn't. The project seems doomed because the processes surround implementing a system just weren't there to the extent they needed to be. What's interesting is that is the same thing that happened to the VA last year.