Should We Stick To What We Know?

The Harris Country Toll Road Authority (HCTRA) operates the toll roads where I live (near Houston, TX) and they do a pretty good job of it.  They upgrade the roads quickly and construction is done relatively quickly.

Since I drive on the toll roads almost daily one of the great things that I like is the EZ Tag system.  This is a RFID tag that scans my account number automatically as I go through the toll booth and charges my account without having to stop at the mainline toll booths.  As a result I zip through at traffic speed (which in Houston varies from 2 mph to 70 mph).  There are a number of toll roads across the country that use this and if you drive with any frequency on them you know what a great advantage they are.

Recently HCTRA expanded the use of EZ Tags to allow me to pay for parking at IAH - Bush Intercontinental Airport.  There is no charge for this.  I just pull up to the entrance and instead of getting a ticket, a scanner reads my EZ Tag and opens the gate.  When I leave I go through another EZ Tag scanner and exit while the system automatically calculates the parking fee and adds it to my account.  At least that's the theory.  Unfortunately reality is much different.

The simple truth is that the EZ Tag system for airport parking doesn't work very well. When it was first installed the gates wouldn't open.   Most recently the gates opened but the transaction wasn't registering correctly.  After 2-1/2 weeks of waiting for the receipt for parking during a recent business trip to show up in my account online I called.  The agent allowed as they had some problems but were "working them out".  After 3 days of investigation they were finally able to produce a receipt.

What had been intended to be a convenience for their customers has turned out to have just the opposite outcome.  I'll be canceling my setup for airport parking, go back to the old pay as you leave routine and just stick with the EZ Tag for the tollway.

Although this wouldn't seem like it would be all that different than the normal transactions apparently it is different enough to cause a problem for HCTRA.  Instead of recording a specific transaction with each scan they now have to combine to scans and calculate the proper transaction.

It isn't unusual for IT to be involved in expanding systems into newer but slightly different uses.  As this example points out sometimes a seemingly simple change turns out to be much more complicated than expected.  The safe reaction is to simply say "Let's stick with what we know".  However, you can't move your business forward that way.

When expanding the capability of a system to do something new it is easy to fall into the trap of saying - "It's basically the same as what we are currently doing, we just have to tweak this one part."  As the EZ Tag situation shows sometimes things aren't as easy as they seem. 

A better way to look at this is to ask yourself how would I do this if I was developing the whole system (with both current and new functionality) from the start.  Then compare this to the tweak approach.  Sometimes you may find the tweak is the full answer and you're good to go.  However you just might find weaknesses or things you hadn't otherwise thought about in the "tweak".

So I'd say don't always just stick to what you know but go in with you eyes open to potential differences in the approach to adding new functionality and you'll have more successful applications and avoid the EZ Tag snafu for a "minor" change.

What are your thoughts?  How do you approach these kinds of changes?

If this topic was of interest, you might also like these:

• Are Your Execution and Delivery Processes Ready for the Internet?
• Are We Creating Problems By Design?
• Are You Ready To Throw Away Your Safety Glasses?
• Or the posts in the "Application Development" category