Penny per Customer Cost for Disaster Mitigation Too Much for Bells

Jeff Pulver and I filed our response to the negative comments which the Bells have unanimously filed with the FCC on our petition, which requires that they do a better job of disaster mitigation this hurricane season and in other disasters than they did last year after Katrina.

The Bells (AKA ILECs) whined that their networks might not be able to route calls out of stricken areas to voicemail or call forwarding and that voicemail facilities themselves might go down in a flood. I’m sure that answer came from their lawyers and not their engineers who must have hid their heads in embarrassment.  They also said that the proposed disaster relief would be too expensive.

To answer “we can’t do it” we enlisted the expert volunteer help of Aswath Rao who understands POTS (Plain Old Telephone System) as well as anyone and better than most.  Aswath contributed the wisdom that the correct answer using current POTS technology is not to route the calls into a stricken area in the first place and to locate backup voicemail facilities in several places around the country.  Please don’t say “duh”, these are our phone companies.  Here’s our lawyer-crafted response with Aswath’s technical guidance:

“The ILEC systems currently have the ability to detect a system overload and re-route or block traffic early during a call, as shown by AT&T’s call blocking in New Yorkon 9/11.  Using this technology, a carrier can provide the emergency voicemail service proposed by Petitioners.  When an emergency or disaster-related system outage (or overload) is detected, instead of “call-gapping” (or blocking), the network would forward the call to a designated location where voicemail services would be stored.  The ILECs could designate, say, 2-3 specific network locations within each of their regions (but geographically distant so as to not be likely to go down at the same time during a physical disaster) where they would invest in spare capacity in order to provide emergency voicemail.  This would alleviate the need to obtain excess capacity in every facility throughout the country.  Only those facilities designated as ones for disaster recovery would need to be upgraded, thereby dramatically reducing the already minimal storage costs.  Furthermore, a carrier could provide voicemail services to customers in this way even if the facilities actually serving those customers were down.  There is no technical requirement that the voicemail service reside on the facilities closest to and serving the customer.  In fact, voicemail services may be provided to customers from their own provider or a totally separate provider with facilities located nearby or somewhere across the country.  Thus, if an ILEC does not wish to host the emergency voicemail services anywhere on their own network or believes their own personnel would be better utilized in a disaster to repair their own networks rather than manage the voicemail service, then the carrier could contract with a third party now to provide the service when needed.”

But what about the cost of all that storage for voicemail?  I used to be a business guy so I worked on that.  Decided to give the ILECs the benefit of the doubt and got the cost of a 250gig harddrive from Tiger Direct.  It’s $120 if you don’t take into account the $70 rebate but maybe the ILECS would forget to send in the UPC from the boxes so ignored the rebate:-}. From our filing:

“Rather than focusing merely at anecdotal indications of the minimal costs, though, Petitioners have attempted to quantify the costs of additional storage necessary to provide emergency voicemail services.  Although the exact configuration for storing voicemail and customer account information would be up the individual carriers, some bounds can be put on the costs involved by looking at the retail price of disk storage (although individual carriers would likely be able to obtain much lower prices due to volume discounts).  For example, a 250 gigabyte Seagate hard-drive is currently available through Tiger Direct for $120, resulting in a price per megabyte of additional storage of $.00048.  …

“To actually provide the service, additional capacity must also be obtained and allocated for customers to record outgoing announcements and receive voice mail messages stored during an emergency.  Petitioners contend that ten megabytes of storage per customer is more than adequate for this emergency voicemail service, resulting in a cost (with retail prices) of approximately $.0048 per customer.  Although a provider may allocate more than ten megabytes for customer subscribers, ten megabytes should be sufficient to store a reasonable amount of messages before the mailbox becomes “full” if it is not accessed and cleared.  Even if we assume a doubling of these costs to allow for additional cabling, controllers, and other equipment, the total cost does not reach even one cent per customer.”

You technical readers are going to complain that I over-engineered this system by ignoring the fact that some customers already have voice mail and by providing maximum storage for every customer even though only a fraction of customers will be affected by any one disaster and even though most customers won’t use the maximum and storage can be allocated dynamically.  Well, at a penny per customer, I didn’t want to cut any corners. Seriously, the point is that storage costs are negligible under the most generous of assumptions.

Let’s look at some practical illustrations of the low cost of storage.  They’re all over the web: free email storage from Microsoft, Google, Yahoo and others.  Free photo storage from Flick and others.  Very much to the point, a new company called GrandCentral is offering free voicemail to the homeless in San Francisco and plans to roll that out nationwide.  Not at all BTW, GrandCentral founder Craig Walker told me that GrandCentral would be willing to provide free voicemail for displaced people in an emergency.  The Bells may want to contact him if they find that a penny a customer of one time cost is too much to pay to make sure their customers stay reachable the next time their physical phone lines are inoperable or unreachable.

Obviously, a penny per customer isn’t the real obstacle to the Bells.  To be charitable, I think these vast bureaucracies are equally resistant to all change.  Of course, they may also be a little embarrassed by the low cost to them of the voicemail services for which they charge a premium of several dollars a month.

It is important that the FCC act as the next hurricane season approaches. There is nothing more certain that future disasters.

I posted less lawyerly rebuttals to BellSouth here and to the Bells in general here.  A general explanation of what the petition calls for and why we posted it is here.

There is a very helpful article In a Crisis, Will Your Phone Work? on CBS News.

Comments

[Galeal Zino]

The technical problem is probably the least complex of all the involved issues to resolve, and the solution you outlined certainly works well. I think there is also another solution that would also work well, although someone more familiar with local number portability (LNP) specifics may want to comment, as I'm only familiar with LNP at a high-level.

In the US, LNP requires each n-1 carrier (the n-1 carrier is the carrier that delivers the phone call to the carrier (n) that services the called party) to dip into a centralized LNP database during the routing process of every phone call to find out if the called party has ported his phone number to a different/new carrier. If the number has been ported, then the result of the database dip is information on where to route the phone call in order to deliver it to the new carrier (the carrier that the number has been ported to). I believe this LNP solution could also be used in an emergency.

Instead of indicating that the phone number has been ported to a new carrier, the database could indicate that the number is ported to an emergency center. There could be a few such emergency centers, distributed throughout the US, operated/funded collectively by the LECs. Each individual LEC would only have to update the LNP database, indicating that the emergency-affected number ranges have been temporarily ported to an emergency center. This should be a single transaction type of process; a process (usually via LNP service providers that operate the LNP databases) that the carriers already use when their customers port their numbers elsewhere. Benefits for the LEC and for everyone include:

(a) This would reduce LEC costs even further as the costs would be split amongst multiple carriers (each LEC wouldn't have to operate their own centers).
(b) This would reduce the reliance on each individual LEC to properly, quickly, and consistently implement call gapping type routing changes on their switches.
(c) The phone calls will never make it to the LEC's network, saving the LECs from network congestion that happens during large emergency situations.
(d) With a few standardized, distributed call centers, each telecom customer would only have one system to learn, no matter where we move to in the US. For example, each customer could get a permanent, life-long emergency voice mail access number etc., and this system would cover all phone numbers (landline, mobile, business, etc.) with a single emergency number.
(e) Restoration to normal could be as simple as the LNP operators removing the temporary port status for the affected numbers...no work for the LEC to undo anything.

A possible downside is reliance on the n-1 carrier to do the LNP dip. Although I believe it is required, there is cost to it so perhaps not all carriers are in compliance.