In: Computer Science
The U.S. Census Bureau signed a $600 million contract with Harris corporation in 2006 to build 500,000 devices, but still weren’t sure which features they wanted included in the units. Census officials did not specify the testing process to measure the performance of the handheld devices. Four hundred change requests were added to the project requirements. Two years and hundreds of millions of taxpayer dollars later, the handhelds were far too slow and unreliable to be used for the 2010 U.S. census.
The above scenario clearly show us the poor planning in requirement analysis and testing the handheld device. Below are the steps they might consider earlier to avoid this situation.
Step 1: Both Census Bureau (CB) and Harris Corporation (HC), should have finalize the complete requirement analysis on what key feature is needed for this handheld device, before even starting the production.
Step 2: The CB board should have to come up with dedicated plans to test the device for durability and reliability. Before putting the devices for production, it is necessary to have multiple stress testing to find the maximum utilization ability of the device.
Step 3: Even after having multiple change request, they are not sure about what features need to be added and how to test it, there should be some milestone wise discussion and testing on the production to get clear idea on how to proceed further.
Step 4: At least by end of two years, when the devices facing slowness, BC should have contact HC and asked for a firmware update to fix the early bugs. Since they have not did that earlier the devices was not at all useful for 2010 census.
Since the project was of heavy budget and number of unit manufactured is very high, we need to be very much careful in each phase, need to have a regular meeting to know the status and to learn more about the project. Even though technology is playing a key role, we need to focus on the Project management side too for any kind of project.