Whenever we use technology, we should always do a pilot. Do you recall IBA's entry test fiasco of medical colleges of Punjab in late 1990s which was widely covered in the media of that time? I saw the fiasco unfolding in front of my eyes from close at IBA where the untested scanner was put on full scale deployment without pilot. This happened despite my protest, and my eventual formal letter a few months before they were going to go full scale. This led to me being removed from the implementation team (Thanks God I was spared the ignominy on media). I have a case study somewhere.
The idea about the importance of pilot is simple, don't put all eggs in one basket and hope upon hope that Murphy's law will not apply. At IoBM, we did a pilot implementation with Smartz CMS. We first implemented in Summer 2018 only for about 500 students before going for a full scale roll out in Fall 2018 with all of our 5000+ students. The consultants for PeopleSoft CMS implementation in 2013 did not do a pilot, and their ensuing full-scale implementation was disastrous with system going down, registrations being rolled back, and students getting agitated. This forced the implementers to install of quick fixes, and hurriedly do several customizations, many of which further destabilized the PS configuration of CMS. The patches so made, were so deep rooted that they could never recover from those disastrous early decisions and the project was shelved after 5 years of huge effort and turmoil, and wasted of investment of around 60million of direct costs, not including the indirect overhead of wasted admin, faculty and student time.
The idea about the importance of pilot is simple, don't put all eggs in one basket and hope upon hope that Murphy's law will not apply. At IoBM, we did a pilot implementation with Smartz CMS. We first implemented in Summer 2018 only for about 500 students before going for a full scale roll out in Fall 2018 with all of our 5000+ students. The consultants for PeopleSoft CMS implementation in 2013 did not do a pilot, and their ensuing full-scale implementation was disastrous with system going down, registrations being rolled back, and students getting agitated. This forced the implementers to install of quick fixes, and hurriedly do several customizations, many of which further destabilized the PS configuration of CMS. The patches so made, were so deep rooted that they could never recover from those disastrous early decisions and the project was shelved after 5 years of huge effort and turmoil, and wasted of investment of around 60million of direct costs, not including the indirect overhead of wasted admin, faculty and student time.
We also did a similar pilot with LMS. We started in Fall 2019 and when the lockdown occurred we had only 50 faculty members with around 100 sections on LMS. The pilot was already working, hence our full blast implementation of all the sections and all the faculty succeeded (but we did go through teething problems of scaling up which required upgrading to cloud). Had it not been an emergency, I would have never allowed it to scale up as fast, and would have scaled it up in a much more phased approach. We also did a pilot of 2nd Hourly and missed exams of the First Hourly, before going through the full blast Final Exams. That too with phased schedules. I am basically a techy person. But, you may have noticed that I pay a great respect to Mr Murphy and his Murphy's Law.
The idea is to prepare for the worst, and hope for the best. But, do it incrementally and with deep introspection of the long term trade offs.
With this preamble, I can now explain why PJET was chosen to go in first on the new system of JMS. It was my decision. It was a pilot implementation. PJET had the least to lose. Others could not have withstood hiccups if any. That done, the rest of the journals can now go on.
With this preamble, I can now explain why PJET was chosen to go in first on the new system of JMS. It was my decision. It was a pilot implementation. PJET had the least to lose. Others could not have withstood hiccups if any. That done, the rest of the journals can now go on.
No comments:
Post a Comment