[Federal Register Volume 59, Number 98 (Monday, May 23, 1994)]
[Unknown Section]
[Page 0]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 94-12525]


[[Page Unknown]]

[Federal Register: May 23, 1994]


=======================================================================
-----------------------------------------------------------------------

POSTAL SERVICE

39 CFR PART 111

 

Presort Accuracy Validation and Evaluation (PAVE)

AGENCY: Postal Service.

ACTION: Notice of proposed program.

-----------------------------------------------------------------------

SUMMARY: The U.S. Postal Service (USPS) published proposed requirements 
for a Vendor Presort Software Validation Program in the Federal 
Register on April 14, 1992 (57 FR 12893-12901). After receiving written 
comments in response to the proposal, the Postal Service decided to 
provide additional opportunity for comment. To facilitate the receipt 
of additional comments, a public meeting was announced in the Federal 
Register on July 21, 1992 (57 FR 32188), and held on August 5, 1992, at 
the USPS National Customer Support Center. As a result of comments 
received, a withdrawal of the proposed rule was subsequently published 
in the Federal Register on September 11, 1992 (57 FR 41716), in order 
to publish the following new proposed rule for a Postal Service testing 
and certification program of presort software that it intends to name 
Presort Accuracy Validation and Evaluation (PAVE).
    PAVE, if adopted, would be a voluntary program in which the Postal 
Service would, upon request, provide testing for certain categories of 
presort software and hardware products to determine their accuracy in 
sorting address information according to the mailing standards of the 
Domestic Mail Manual (DMM). This program would assure those using 
presort software packages that a PAVE-certified product, if used 
properly, would have the capability of performing its intended function 
according to the current mailing standards of the Postal Service.
    Because make-up errors in mail preparation cannot be eliminated via 
software alone, the use of a PAVE-certified presort product to produce 
a mailing would not assure rate eligibility. As part of this program, 
manufacturers of certified presort software would need to inform users 
of their products that proper use of presort software (e.g., selecting 
appropriate options for a given sortation, using the correct parameter 
settings, etc.) would remain critical to ensuring accurate sortation.
    The Postal Service is seeking comments from interested parties, 
including developers and users of presort products, to ensure the most 
effective design and implementation of this proposed program.

DATES: Comments must be received on or before June 15, 1994.

ADDRESSES: Written comments should be mailed or delivered to the 
Manager, Business Mail Acceptance, 475 L'Enfant Plaza SW., room 8430, 
Washington, DC 20260-6808. Copies of all written comments will be 
available for inspection and photocopying between 9 a.m. and 4 p.m., 
Monday through Friday at the above address.

FOR FURTHER INFORMATION CONTACT: George T. Hurst, (202) 268-5232, or 
Lynn Martin (202) 268-5176.

SUPPLEMENTARY INFORMATION: Many mailers rely on computer software to 
generate mailings for postage discounts. Two areas where computer 
software has proved particularly advantageous in mail production are 
the matching and coding of individual address records and the 
presorting of address lists. To validate the accuracy of address 
matching software, the Postal Service has established the Coding 
Accuracy Support System (CASS), which provides a common platform to 
measure the quality of this software. The Postal Service proposes, with 
this notice, the implementation of a similar program to validate the 
accuracy of presort software products in specific presort categories. 
This program would be voluntary and available to all developers of 
presort products that sort address information according to the presort 
standards of the DMM. It would be available to test products configured 
for personal, mid-range, or main frame computers including mail-sorting 
optical character readers. The Postal Service does not propose that the 
use of PAVE certified software be mandatory for obtaining presort 
postage discounts.
    For purposes of PAVE certification, eligible presort product 
developers are defined as those firms that initially develop a presort 
product, regardless of commercial availability. This definition also 
includes those that make modifications that alter the sorting logic, 
documentation generation capability, or add to or revise the original 
parameters of a software product.
    A presort product is defined as a complete set of computer program 
modules or equipment that presorts address lists or mailpieces 
according to DMM standards for one or more types of mailings.

Program Overview

    Developers of presort software and/or hardware would apply for PAVE 
certification of their products. On receipt of the application, the 
Postal Service would forward the appropriate test file of address 
records to the developer. The developer would sort the file with its 
presort product and print documentation supporting accurate sortation 
and rate application (along with other documentation printed by the 
software). This information, along with the original media, would be 
returned for evaluation to the USPS National Customer Support Center 
(NCSC) within a specified time frame.
    The Postal Service would then evaluate the documentation for 
presort accuracy. In addition, the Postal Service would evaluate the 
accuracy, format, and content of; required postal documentation (where 
applicable), facsimile mailing statements (if generated), and any other 
postal forms generated by the software product in the category being 
tested.
    If the Postal Service determines that the presort product satisfies 
the current presort standards, the developer would be issued a PAVE 
certificate for that product in that category. The certification would 
be good for 1 year, or until the next PAVE cycle. PAVE testing would be 
conducted from August through December, as detailed further in this 
notice. A list of PAVE-certified software/hardware developers would 
then be published in the Postal Bulletin (or a similar publication/
listing as is currently used for CASS/MASS certified vendors).

Selecting Specific Presort Categories

    The Postal Service would not be able to test individually all 
presort categories every year. The Postal Service would choose a 
limited number of presort categories annually for PAVE testing. Presort 
product developers would be notified of the new test categories as well 
as any changes made to previous tests to be evaluated in the next PAVE 
cycle approximately 3 months in advance of actual testing, so that 
developers could prepare for and order the appropriate test files.
    Presort categories would be chosen based on, but not limited to, 
the following criteria:
     Presort categories most often used by different mailers 
(to ensure testing of the most commonly used products).
     Presort categories where significant mail preparation 
error is being detected by the Postal Service.
     Presort categories that are relatively new.
    For the first PAVE test cycle, the Postal Service proposes the 
following domestic presort categories:
     Presorted First-Class.
     First-/third-class barcoded letter-size:

--Tray-based.
--2-Tier package-based.
--3-Tier package-based.

     Second-class presort (carrier route, 3/5 digit, and 
basic).
     Third-class presort (3/5 digit and basic).
     Third-class carrier route presort.
    A presort product developer requesting one or more of the files for 
PAVE testing would also receive a technical guide that details 
specifications for each test category. These specifications would 
include such information as:
     Mailpiece specifications (mailpiece thickness, weight, 
length, and height).
     Mail processing category.
     Maximum pieces to a tray (if applicable).
     Minimum pieces to a tray (if 3/4 full tray is a pertinent 
criteria).
     Mail entry point.
     Sortation levels allowed (if applicable).
     Overflow trays (if allowed).
     Total number of records (addresses) contained in the file.

OCR Testing

    As detailed later, all the test files would be provided to 
requesting presort product developers in electronic media (magnetic 
tape, cartridge, and diskette). However, for optical character reader 
(OCR) applicants, the Postal Service would provide the First-/third-
class barcoded letter-size test(s) as a physical test deck of actual 
mailpieces. The Postal Service proposes that the physical test deck, 
rather than an electronic file, would better accommodate the actual 
sortation process of an optical character reader.
    This physical test deck would be similar to that used in MLOCR 
Accuracy Support System (MASS) testing, consisting of several thousand 
sample mailpieces that would be sent to the applicant's location for 
OCR presorting. The entire test deck would be presorted as would a 
regular mailing in that particular category, trayed, labeled, and 
documented, and then returned to the Postal Service for PAVE 
evaluation.
    Similar to other PAVE participants, eligible OCR participants would 
consist of those manufacturers that initially developed the presort 
software product used in the OCR, regardless of commercial 
availability, or those that made modifications that alter the sorting 
logic, documentation generation capability, or add to or revise the 
software's original parameters.

Test Cycle and Frequency

    The Postal Service proposes the following PAVE test cycle for 1994:
1. Submitting Applications (August 1-31)
    Software developers would complete and forward order forms for a 
specific test file(s) to the Postal Service for receipt no later than 
August 31.
2. Distributing of Test Files (September 1-30)
    NCSC would mail the test file(s) to requesters during this period.
3. Returning Test Files (September 1-October 15)
    Presort product developers would process the test file through the 
presort product being tested and return the sorted file, along with 
hard copy documentation supporting accurate sortation and rate 
application, for receipt by NCSC no later than October 15.
4. Evaluating and Responding (November 15-December 15)
    The Postal Service would examine the hard copy documentation for 
accuracy and provide evaluation results to the presort product 
developer no later than December 15.
5. Retesting (November 15-December 31)
    Developers would have this period to request, process, and be 
reevaluated on those test(s) that they had failed in the initial test 
cycle.
6. Publishing PAVE-Certified Developers (January, 1995)
    A list of PAVE-certified presort product developers, product names, 
version numbers, and presort categories would be published in the 
Postal Bulletin (or a similar publication/listing as is currently used 
for CASS/MASS certified vendors).

    Note: Although the Postal Service proposes the above PAVE cycle 
to establish reasonable time frames and cut-off dates for each 
process, steps 2 through 5 would be expedited where possible.

Application Process

    To apply for PAVE certification, eligible presort product 
developers would complete the order form shown as Exhibit A of this 
notice. Only one form would be necessary regardless of the number of 
presort categories being applied for. The completed form would be sent 
to the following address: Pave Certification Program, National Customer 
Support Center, United States Postal Service, 6060 Primacy Pky Ste 101, 
Memphis, TN 38188-0001.

PAVE Test Files

    PAVE test files would vary in record quantity and ZIP Code variety 
depending on the presort category being tested. However, all the 
address records in the files would consist of randomly selected primary 
street records appropriate for the ZIP Code, ZIP+4 code range, or 
delivery point code numerics used (the correction character numerics 
would not be included). This use of error-free addresses would 
eliminate problems arising from presort products that incorporate some 
address verification mechanisms prior to sortation.
    The test files would be configured in the following media:
     Magnetic Tape:

--6250 BPI EBCDIC.
--6250 BPI ASCII.
--1600 BPI EBCDIC.
--1600 BPI ASCII.

     Cartridge:

--IBM 3480 38K.

     Diskette:

--3-\1/2\'', 1.44 Megabyte MS-DOS.
--5-\1/4\'', 1.2 Megabyte MS-DOS.

     Physical Test Deck (for OCR software/hardware developers 
only)

--Standard #10 white window envelopes (4-\1/8\'' x 9-\1/2\'') made of 
24 pound basis weight paper stock with inserts preprinted with OCR-
readable addresses and POSTNET barcodes.

Processing Test Files and Providing Documentation

    Once received, the developer would sort the test file(s) at its 
location with its presort product according to the presort category 
being examined. The developer would generate hard copy documentation to 
support the accurate sortation and rate application of the addresses on 
the test file. This would include such information as the following:
     Documentation required to accompany the mailing statement 
(if applicable).
     Tray/package audit trails.
     Tray, sack, label generation (samples of each type).
     Summary reports.
     Parameter reports (parameters used to run the test file).
    In addition, the developer would generate (as an output of this 
test run) and submit for PAVE evaluation other forms of documentation 
available to users of the product, such as:
     Mailing statement facsimiles (if mailing statement 
facsimiles are not generated by the software product, the manufacturer 
would have to provide rate qualification totals for PAVE 
certification).
     Other postal forms facsimiles.
    At previous commenters' requests, the Postal Service considered the 
possibility of having developers send their products to the Postal 
Service for PAVE testing rather than providing test files to be 
processed at developers' locations. After initial review, this was felt 
to be inefficient or inappropriate for the following reasons:
     Various operating platforms/systems might not be available 
for the Postal Service to run all the products.
     Postal Service examiners would likely require in-depth 
training on many products in order to evaluate their capabilities 
properly.
     Tests might become too subjective given examiner's varying 
knowledge of different products.
     Security and liability issues might arise given the 
receipt and usage of presort software programs not otherwise available 
to the general public or in advance of their commercial release.

Evaluating PAVE Tests

    Evaluating a returned test would consist of the Postal Service 
examining the required hard copy reports returned by the presort 
product developer and, for OCR evaluation, the physical test deck. This 
evaluation would focus primarily on proper sortation according to DMM 
standards but would also include an examination of the content and 
clarity of supporting documentation (required to accompany mailing 
statements), the content, format, and clarity of facsimile postal 
forms, as well as the accuracy of other computer-generated 
documentation submitted. To be PAVE-certified, the examination would 
have to conclude that the results were in complete accord with the DMM 
standards that were applicable at the time of the test.
    In many cases DMM presort standards allow for a range of correct 
answers to a specific sortation. As an example, the precise number of 
pieces to a tray or sack is not mandated by standard. Changing this 
parameter can create different mailings from the same address list and 
yet still satisfy DMM standards. For this reason, no one specific (or 
optimum) answer would be mandated under PAVE certification unless only 
one answer was correct on a specific issue.

PAVE Certification

    PAVE certification would be valid for 1 year or until the next PAVE 
cycle. For those choosing to test during the normal test cycle, PAVE 
certification would be valid from January to January. Upon successful 
test evaluation, developers would be contacted by the Postal Service in 
writing and issued a PAVE certificate noting the specific software 
tested, the categories tested, the date of certification, and the date 
of certification expiration.
    A list of all PAVE-certified presort product developers would be 
published in the Postal Bulletin or a similar publication/listing as is 
currently used for CASS/MASS-certified listings. This list would 
include the developers name, product name, version numbers, certified 
presort categories, and a company contact name and telephone number.
    PAVE certification would not remain valid if a presort product 
developer made significant changes to its product within the 12 months 
following certification. Developers initiating such product changes 
would need to apply for certification for this new, changed product. A 
significant change would be a key alteration of the product's basic 
sortation logic; a major change in the content, layout, format, or 
availability of computer generated documentation or facsimiles, or a 
modification that results in significant differences in software 
operator use.
    Some software changes would not be significant enough to require 
recertification. PAVE-certified presort product developers would be 
required to notify the Postal Service whenever changes were made to 
their certified products to determine whether recertification were 
necessary. To aid in the quick release of improved presort products to 
accommodate mailing industry needs, the Postal Service proposes the 
establishment of a toll-free telephone number, to obtain information on 
whether the changes would warrant recertification.

DMM-Initiated PAVE Cycle

    Any significant change in the make-up and preparation standards for 
presorting might intersect the normal 12-month PAVE certification 
period. Such a change might require a presort product developer to 
modify its product enough to trigger a recertification or a DMM-
initiated PAVE cycle. To provide time for recertification, the Postal 
Service would attempt to delay implementation of significant presort 
standard changes (those deemed significant enough to require 
recertification) by 120 days whenever practicable. The Postal Service 
would implement these changes 120 days after the final date of 
notification (i.e., the date of publication of the final Federal 
Register rule) whenever possible. (Some changes, such as those 
resulting from congressional action or postal rate commission 
proceedings, might require more immediate implementation.)
    If the Postal Service conducted PAVE testing out of the normal 
cycle to accommodate such DMM changes, presort products tested and 
approved during this time would maintain certification for 1 year 
beyond the next normal test period. For instance, if a DMM change 
became effective May 1995, a presort product that became PAVE-certified 
in July 1995 (out of cycle to meet the new standard) would be issued a 
PAVE certificate valid from July 1995 until January 1997 (unless 
further DMM changes warranted significant software updating).

Out-of-Cycle Fees

    Participating in the PAVE certification program during the normal 
testing cycle would be free of charge. In those instances when a DMM-
initiated PAVE cycle was conducted, participation would also be free of 
charge. A fee, however, would be charged for those participating in the 
PAVE program outside the normal test cycle. This fee would be assessed 
as follows:

Test File--Initial presort category (file generation, postage and 
handling)--$50.00
    ($10.00 for each additional test category requested at the same 
time)
Certification--Results evaluation (presort analysis and documentation 
review)--$200.00
    ($25.00 for each additional test category submitted at the same 
time)
Minimum Fee--$250.00

    Out-of-cycle PAVE certifications would expire at the end of the 
next normal test cycle (e.g., a certification obtained in June would 
expire December 31 of that same year).

PAVE Test Failures

    Participants would receive notification of errors during PAVE test 
evaluations so that retesting could be accomplished within the normal 
PAVE cycle. Although November 15 through December 31 is proposed as an 
established time frame for retesting, the Postal Service would strive 
to expedite the mailing and evaluation of retests whenever possible. 
The first 2 attempts for a specific category would be free of charge. 
All subsequent certification attempts would be charged the out-of-cycle 
fees.
    Software not achieving certification prior to December 31 would not 
be included in the list of PAVE-certified presort product developers to 
be published in January of each year.

Standardization of Testing Data

    Based on a variety of previous comments, the Postal Service would 
not mandate standardization of presort software documentation, layout, 
and format with this proposal. The mailing documentation currently 
generated by presort software would be evaluated to ensure that all 
data necessary to support the rates claimed in the mailing category 
tested were present and could be readily identified and understood.
    The Postal Service believes, however, that eventually a more common 
framework for test results needs to be established to have an 
efficient, accurate PAVE review and certification process.
    The Postal Service proposes to explore the feasibility of having 
software manufacturers return an electronic file rather than hard copy 
documentation for future PAVE testing. From this electronic file, the 
Postal Service could perform computer analysis of PAVE test results, 
begin to develop analysis software for a computer grading solution, or 
print documentation in a common format for ease of review. In order to 
pursue this concept, the Postal Service would need to establish a 
common electronic file structure with standardized data fields.
    Although the issue of using an electronic file is not specifically 
part of this proposed rule (hard copy documentation is proposed for 
manual evaluation for initial PAVE tests), comments concerning its 
eventual implementation are requested. A potential file structure has 
been drafted and is available on request to those wishing to comment on 
its content and layout.

BILLING CODE 7710-12-P

TP23MY94.000


TP23MY94.001


BILLING CODE 7710-12-C
Stanley F. Mires,
Chief Counsel, Legislative.
[FR Doc. 94-12525 Filed 5-20-94; 8:45 am]
BILLING CODE 7710-12-P