Sunday, May 23, 2010

NTTA Put to the Test...

This last Friday (May 21st, 2010) I had to make a trip across the DFW metroplex.  I drove on the George Bush Turnpike (GBT) to see how long that it would take before the North Texas Tollway Authority (NTTA) sent me the first bill for my use of the GBT as well as to see how many bills that I would receive because I passed though several automated Electronic Toll Collection (ETC) booths.  This test is for a car that I know they have in their system.  Theoretically, I should get a bill within a month or so.  We will see.

Last month I took a new vehicle on the GBT taking the same path as last week's journey.  Because that was a new vehicle, I don't expect to receive a bill for at least a few years.  That way the NTTA can charge me huge late fees and administrative fees in the process.

In my previous blog post regarding the NTTA , I suggested that with the way they handle (or perhaps better stated don't handle) sending out bills and reminders in a timely manner. In my case from my previous blog post, it took over two years before they finally sent me my first bill via a collection agency. When I received the bill, I was charged $175 for $7.34 worth of toll charges. I suggested in that blog post that if the NTTA only treated 20% of the 500,000 (source: CBS estimate) daily users of the NTTA toll system that they could earn approximately $2.3 million per day yielding a total annual income from excessive fees of over $800 million.

This made wonder how much that the NTTA actually earned last year. So scanned through their 2009 NTTA Summary Annual Report. I was shocked by the overview of their financials and the accounting of the use of the tollway at the end of the report.  For reference, here is a screen capture of that information.

The NTTA reported that they only had 456,700,028 transactions for all of 2009. I presume that a transaction is each individual toll that a driver passes through.  If this is the case, then in 2009 the NTTA earned approximately $.73 per transaction.  The NTTA used to charge around $.75/transaction on the Dallas North Toll Road.  So, the total number is relatively in alignment with what they should have made with a small amount of loss.  However, in 2009 they did raise toll rates.  The NTTA clearly didn't break even on their toll/transaction rate for 2009.

I wonder how much of the revenue was generated by late fees and administrative fees versus timely paying Toll Tag and non-Toll Tag customers.

If the NTTA wants to offer true financial accountability in order to make their case of being robbed by non-paying users of the toll roads, they should include in their annual report a full breakdown of transactional types (e.g. Toll Tag, non-Toll Tag, paying, non-paying, and the tolls, fees and late payments recouped by those transactions. This is the only way to truly make their case in a convincing way.

As a side note, there were several noteworthy quotes from the 2009 NTTA Annual Summary Report that I found quite interesting.  Here is the first quote.

After six months of all-ETC operation in 2009, traffic was meeting revised projections. Revenue continued to increase as more ZipCash invoices were sent and paid for travel on the PGBT.  In fact, revenues exceeded projections.

Lets look at the those actuals.  First, here are the number of transactions for 2009.
The number of transactions per month has some minor fluctuation but largely similar.  Roughly 35 million transactions per month.  Now lets look at the revenues for 2009.
I find it interesting that although the number of monthly transactions didn't change that much throughout the year but in the months of September and November the revenue was almost double the preceding months.  I wonder if that is when the NTTA started collecting late fees and administrative fees on "non-paying" customers like me who never received a bill until I was sent a collections notice.  That would certainly explain why the NTTA "revenues exceeded projections".

On to the next savory quote.  This one focuses on their great customer support.  Before you read this quote note that when I called in, I waited over 30 minutes before anyone picked up the phone.
In 2009, the Roadway Customer Service team handled 17,985 incidents, representing an average of 346 per week, with an average response time of less than 12 minutes.
I'm not sure if the NTTA is aware of this or not but an average response time (e.g. the amount of time you wait on hold for someone to answer the phone) of nearly 12 minutes is INSANELY too long to promote a happy customer.  If the NTTA only works a 40 hour week, that means that they took less than 9 calls per hour.  The NTTA either has really long calls per agent or have a very lax requirement to pick up the phone in a timely manner.

Note also from the same report that in 2009 the NTTA had to reorganize 170 employees and let 19 go.  Not sure if the 19 were from customer service or not but that might have contributed to the long hold times.

Here is the next quote.  Not so savory.
Fine-tuning our ZipCash invoices As with any ambitious new program, our ZipCash launch required some fine-tuning. Some ZipCash customers, having never received an invoice from the NTTA before, were initially confused about the process for billing and payment. We gave our invoices and notices a user-friendly makeover to help customers distinguish each step in the payment and collections process.  To minimize confusion about paying ZipCash tolls, each phase of the new three-step invoice process—ZipCash, ZipCash Late Notice and Violation Invoice— has a distinct look. We hope to encourage motorists to pay for the use of the toll road before any late or administrative fees are applied to the transactions. We want the customers’ payment process to flow as smoothly as their driving experience.
Glad they made the invoices pretty and all but if they don't send them, it doesn't really matter how pretty they are.  But there is a bonus for NTTA with this strategy.... a lot more late fees and administrative fees to collect.

Here was a quote that would be interesting to know the rest of the story.

The story behind administrative fees It is important to the NTTA that we conduct all of our business in ways that are fair, clear and easily understood.  One way we do that is by working to see that the 92 percent of motorists who pay their tolls are not subsidizing nonpaying motorists’ trips on toll roads.
The NTTA takes a hard-line stance on habitual violators. Our collection process includes an administrative fee, authorized by statute, to help pay for the cost of collection. This is not a stream of additional income but rather a way of recouping some of the costs associated with enforcement.
However, the NTTA realizes that mistakes happen, confusion occurs or invoices are overlooked. For that reason, we waive 67 percent of administrative fees incurred if any violator takes care of the balance before the outstanding invoice is transferred to a collection agency. We want to do all we can to help customers avoid any additional fees.

With respect to the 92% rule, that sounds like the NTTA is saying that 12% is the number of nonpaying motorists.  I wonder if the 67% of waived fees were for the customers like me that called in and demanded that the erroneous fees be dropped... and the NTTA customer service representatives begrudgingly agreed.

The next humorous thing I found in the report is that the NTTA setup a Facebook (FB) page to become more social.  So, I went to FB and searched on NTTA.  Unfortunately I didn't find the official NTTA page.  However, what did show up was quite interesting... NTTA Fines Suck! FB page.  This was a vibrant community of 90 NTTA anti-friends that were all to vocal of their disapproval of abusive NTTA fines.  I had to resort to Google to find the official NTTA FB page.

I can't wait to see the 2010 NTTA Annual Summary Report.  I wonder if the overall revenue will nearly double like it did in the months of September and November of 2009.

I hope you enjoyed this exploration into the the realm of the NTTA and found my commentary interesting.  If I have misrepresented anything or got some fact wrong, please don't hesitate to correct me.   When correcting though, please site all references for everyone's benefit.

That is it for tonight.  Have a great day!


Wednesday, May 12, 2010

MythMobile: MythTV Streaming for Mobile Devices

I am a long time user of MythTV. I am also a recent user of Apple's iPad, iPhone and iPod Touch mobile devices. One of my personal projects has been to figure out how to watch my MythTV recordings on these and other such mobile devices.  For example it should work on Android based devices as well. After much research and trial and error, I have finally come up with a solution that seems to work well.

The solution is relatively straight forward.  I transcode the recordings and store them in a user friendly naming scheme in a folder that is shared via Apache web server and Samba file share.  For extra convenience, I copy the recordings via rsync to an Internet accessible web server so that I can watch my recordings from anywhere.

The biggest challenge to this project was getting the transcoding right. Here is a sample invocation of ffmpeg that works for me.
# nice ffmpeg -y -i "${input_file}" -s 480x352 -r 29.97 -g 300 \
   -vsync 1 -vcodec mpeg4 -b 1280k -maxrate 1536k \
   -async 1 -acodec libfaac -ab 1024 -ar 44100 -ac 2 \
   -flags +aic+mv4 -mbd 2 -cmp 2 -subcmp 2 -bufsize 2M \
The three main video codecs that I recommend are MPEG-4 (e.g. mpeg4), DivX (e.g. libxvid), and H.264 (e.g. libx264).  The best video quality to me was a tie between MPEG-4 and DIVX.  You also can adjust the bitrate/maxbitrate to find the video quality/file size that best meets your needs.  A bitrate/maxrate combination of 1280k/1536k resulted in good video quality and about a 10x reduction in the file size from the native format (e.g. MythTV mpg) to mp4.

To simplify and automate my environment, I wrote two perl scripts that manage creating and presenting the mobile friendly video files.  The first perl script (e.g. transcodes, renames, and stores the video files in /var/www/MythMobile.  Note that Apache Web Server and Samba File Server have been configured to make /var/www/MythMobile available on my local network.  When running this script, you can specify via the optional flags the recordings directory with --r recordings_dir and the web directory with --w www_dir.  If no flags are specified, the default locations are /var/lib/mythtv/recordings for the recordings directory and /var/www/MythMobile for the web directory.

The second perl script (e.g. index.cgi) can be placed into a folder within your web server document root.  For example, I placed it in /var/www/MythMobile on my server.  This script facilitates presenting the video files in a Mobile device friendly format.  Note that I had to modify the Apache web server configuration (e.g. httpd.conf) with the following modifications.
  1. Add index.cgi to the DirectoryIndex.
  2. Add the .cgi handler by adding "AddHandler cgi-script .cgi".
  3. Add ExecCGI to the Options of the containing Directory.
With the installation of the index.cgi script and these three web server changes changes applied, the video files are presented with much more user friendly format as you can see from the pictures below.

Here are the vertical and horizontal views on an Apple iPhone.

Here is the vertical view on an Apple iPad.

Here is the horizontal view on an Apple iPad.

I bundled up the two perl scripts and a README and made them available as a zip file named

Check it out and try it for yourself.