Census Bureau proposing to scrap handheld plan, going back to paper
We do the official census here in the U.S. every 10 years so there’s plenty of off-cycle time for improvements. One of the plans for 2010 is about to be scrapped if the Census Bureau has it’s way. In 2006, the Bureau awarded a contracts for over half-a-million handheld computers to help collect data from folks that didn’t send in their paperwork. Sounded like a great use of mobile tech, but now the Bureau plans to abandon the idea.As I read the story though (warning: 15-year IT background surfacing) it seems to be more of a failure to manage the technology project rather than a failure of the technology itself.
“Since awarding the contract, the project has experienced constant setbacks, including changing system requirements that led to increased costs and missed deadlines.”
Reading further into the article, it appears that 400 new or changed technical requirements were delivered to the project team earlier this year, after a May 2007 test identified “development and scoping problems”. You can have the best technology in the world, but I’ve always said it doesn’t mean a thing if it doesn’t meet the project requirements. Sadly, the cost of not going with the handhelds and reverting back to paper forms will add another $2.2 and $3 billion although some of those costs aren’t related directly to the handhelds. The Bureau still plans to use them for address verification, so don’t expect to find 525,000 handhelds on the cheap at a liquidation house. Thanks for the tip, Roger!
Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
Nobody mentions who got the contract.
Will you?
Being in the software development industry, you comment about “400 new or changed technical requirements were delivered to the project team earlier this year, after a May 2007 test identified “development and scoping problems” strikes a nerve.
If 400 requirements were added or changed, that tells me:
- inadequate requirements definition took place in the initial design phases, which the government both participated in and had to approve
- “scope creep” was introduced by the government (trust me on this one), often found in the testing phase when a group of testers insist a new feature is required
- Development failed to understand the requirement (and why wasn’t this caught during periodic software review by the customer?)
- the end goal the product needed to meet had changed dramatically during development, prompting new requirements (which often happens during long development periods)
- the government demands updated hardware or COTS software, which necessitates coding changes to accomodate
My sympathies lie with the contractor who’s developing the software. Now, I wonder if that’s my company who holds that contract?
Whew, not mine!
“Harris will serve as the systems integrator and provide overall program management. Members of the Harris team include: Accenture LLP, which will provide mobile computing applications and enterprise support systems; Unisys Corporation, which will provide nationwide support and service for approximately 500 Census Bureau field offices; Dell Computer Corporation, which will provide office computing equipment; High Tech Computer Corporation, which will provide the mobile computing equipment; Sprint, which will provide telecommunications services; Oracle, which will provide database support; Client Network Services, Inc., which will provide engineering and field technician support; and Headstrong, which supported the enterprise architecture development for the 2010 Census.”
From http://www.harris.com/view_pressrelease.asp?act=lookup&pr_id=1818