PROJECT MANAGEMENT, PLANNING, SOFTWARE, and TRAINING
CPM (Critical Path Method) was developed by Mauchley Associates for Construction and used detailed work activities and dummy restraints.
PERT (Program Evaluation and Review Technique was developed by Donald Malcom, the Project Manager for Booz Allen & Hamilton Consulting Planning for the U.S. Navy Polaris Missile Program at Lockheed.
The PERT type of planning was required by Aerospace Corporation of Philco-Ford Western Development Lab, Palo Alto, California for electronics manufacturing of On-Board Computers for Satellites, & construction of Tracking and Data Acquisition System sites.
CPM/PERT was required by Hershey Chocolate of Day & Zimmerman for the construction planning of the plant and installation of equipment at the Hershey Chocolate plant in Oakdale, California 1963-65.
CPM/PERT was required by SNPO (Space Nuclear Propulsion Office) of Catalytic Construction in the Nuclear Rocket Development Station construction of test sites and the testing of Nuclear Engines at Jackass Flats, Nevada 1965-67.
CPM/PERT was required by BMI (Basic Metal Inc.) of United Engineers in the plant conversion from Magnesium to Titanium Sponge production in Henderson, Nevada 1967-68. Once a month I would fly from Chicago to Las Vegas and spend a week collecting input and fly back to Chicago and then fly to Minneapolis and renting time at night on a CDC 3600 512kb computer, located at the CDC (Control Data Corporation) Corporate office, running on a tape operating system. I gave a seminar on how to improving Planning and Scheduling techniques and improvements that were needed of the CDC computer program I was using. So many people from my seminar went to their salesman and they went to their Director of Programming. Then they all came to my seminar room where I was still talking with other attendees. The Director said "You have a better understanding of organization needs and we offer you a copy of our source and documentation to improve our product and provide us with a copy of the modified source and documentation. /p>
I developed a training program and trained approx. 600 people for merger planning and scheduling of two railroads in Illinois; the Chicago & Northwestern Railroad and the Milwaukee Road in 1969-70. The program had to be completely rewritten to run on a 128K computer and converting from FORTRAN II to FORTRAN IV. It was redesigned to run on a smaller Computer by reducing 8,000 activities to 7,500 activities. and added Planning and Scheduling advantages not explored by Microsoft Project or any other planning and scheduling network diagramming and computer simulation software existing. The technique was developed using logic, construction, engineering, computer science education and experience.
Originally written in FORTRAN II, rewritten in FORTRAN IV (FORmula TRANslation) of over 12,000 instructions, with 15 main programs, massive common that had to remain in memory the original limit was 8,000 activities and restraints only had to be reduced to 7,500 on a 128 KB ram IBM 360 computer by re-structuring the common blocks. RAM (Random Access Memory) available today should be able to handle 30,000 Work Activities, Events and Restraints.
Actually, using a 32 bit processor, you are limited more by addressing capabilities of the computer. A 64 bit processor The program was original written for a CDC computer with a better instruction set that did not require the use base register addressing as the IBM computer did, so the CDC computer only took 18 minutes to run my test data. When rewritten for the IBM 360 clone, the same test data took 6 hours to run because of the IBM processor instruction set. The CDC program was originally written to look for a known value in 8,000 unknown locations. Base Address calculations of the IBM computer were eating up time without producing, I had to solve the problem cause which was up to 7,500 calculations to find a know value hiding in up to 7,500 unknown locations for each activity.
I used a concept of, if a problem cannot be solved using a normal method, try turning the problem inside out. So I changed the problem inside out, look for a unknown value in a known location. As each activity and event was processed for the first time I assign a dense sequential (pseudo) number in each node record, and changed the program to look for an unknown value in a known location the pseudo number location in common. This saved up to up to 7,499 address calculations per network activity or restraint. By doing this we were able to reduce the computer time from 6 hours to 2-1/2 hours, a reduction to 41.7% of the previous time. Realize computers were not very fast in the 1970's. The computer data processing efficiency today is much better than 1970's technology and will get better in the near future with 64 bit computers.
It is not the speed of the CPU in Megahertz or Gigahertz that determines the output, it is the architecture of the computer, the instruction set, that determines the quantity, from 1 to 8 or more CPU cycles that the CPU will have to perform to complete each different instruction type. With some types of data a slower CPU with the right instruction set will out perform a faster CPU. With speed you generate heat, heat can cause failure or erratic functioning.
ATAD Corporation needs a gigantic (30,000 or more activities) CPM/PERT network to plan and schedule humongous projects to the level of comprehension by the workers that will do the original construction or modification, procurement and installation office furniture, server, client hardware, network design and cabling, software design, programming, test data preparation, program testing, installing, & testing; Executive Office worker personnel interviewing, hiring and training, etc. Without a very detailed level of activity planning of labor, equipment and materials work activities and different types of restraint activities that affect those work activities such as Finance, Cash Flow Analysis; the time required for the procurement of each item, each vendor lead time for procurement, delivery of material and production of items it is very difficult to make a good estimate of the time required to accomplish those tasks We revived the Computer Project Management System I developed from a new concepts of CPM PERT and added my other modifications and called it COPES (Computerized, Organized, Planning, Estimating and Scheduling).