For: Digital talking book distribution analysis. Task 2 - System design (Final report : June 29, 2006)
Section 2 - Current Audio Book Distribution Operations
This section of the report presents information on current audio book distribution operations, a thorough understanding of which is necessary in order to plan the successful implementation of the future Flash Memory-based digital book distribution system. An important aspect of this understanding is knowledge of the variations in audio book distribution operations among the libraries, and to what extent such variations can be accommodated and/or to what (if any) extent and how such operations must be modified, or possibly standardized, in order to facilitate implementation of the new system.
To this end, further analyses were performed in order to both update and augment findings presented in the final Phase 1 report. Site visits were also performed at a sample of network libraries in order to gain a thorough understanding of audio book distribution operations in the network, including the variations therein, with an eye towards what changes will be necessary in order to implement the future system.
2.1 Results of Additional Analyses
Additional data collection, compilation and analyses were performed in order to update and augment the statistical profile of current audio book distribution operations described in the Phase 1 final report. These analyses addressed audio book production, circulation, collections, collection turnover, and library information systems.
2.1.1 Audio Book Production
As stated in the Phase 1 final report, a major assumption made for planning the future audio book distribution system is that production of DTB titles in the future system will be identical or very similar to that for RC books in the current system and during the recent past. That is, the number of new book titles produced per year, the quantities produced (i.e., copies per title, for those titles that are mass-produced, with relative anticipated demand considered), their lengths (i.e., duration in minutes), and the mixture of content offerings will be approximately the same as that for the current system and would not vary whether the Hybrid or All Mass-Duplication system is implemented.
Analyses of RC book production were performed using data from the NLS Production Inventory Control (PICS) system. Source data consisted of all RC records in the system for titles shipped during FY 2001–FY 2005. These analyses resulted in both an update to and correction of results presented in the Phase 1 final report (FY 2005 data were incomplete at the time and therefore not included, and the PICS data extract provided by NLS for use in Phase 1 had a few records missing). The most important results of these analyses are summarized below.
- Titles Shipped: 1,872 RC titles were shipped during FY 2005, and production has averaged 1,987 titles per year over the last five years, as shown in Appendix 1. Production has been at an almost uniform level in recent years. Therefore, although FY 2005 production was slightly below this 5-year average, 2,000 new DTB titles produced per year (in steady-state operations) is still a good estimate to use for planning. Even for the transition years (FY 2008–FY 2011 or FY 2012), 2,000 titles per year is a good estimate to use (although whether all titles will be produced in both RC and DTB formats (in various quantities) in all transition years is being determined in the Transition Plan task of this project after completion of the system design).
- Number of Copies Produced per Title: Results presented in the Phase 1 final report showed that the number of RC copies produced per title varies from lows in the upper 400's to slightly over 2,000 for the titles with the very highest production quantities. As shown in Appendix 1, the most recent 5-year average production lot size was 911 copies per title, which is very slightly below (1.5%) the 925 copies per title value that was used for planning in Phase 1. These values showed little variation among the five years in the time period; the value for FY 2005 (896) is slightly below this average.
- Number of RC Containers: The statistics shown in Appendix 1 agree with findings from Phase 1 of the project, i.e., that approximately 98.5% of RC titles require one container (whether a 4 or 6-cassette type), and only 1.5% of titles require 2 (or, in very rare cases, more) containers (which is equivalent to the average book requiring 1.02 containers).
- Length of Titles: As cited in the Phase 1 final report, RC title durations range from only a few minutes to dozens of hours, but both extremes are unusual. As shown in Appendix 1, the 5-year average length (unweighted by quantity produced) was 637 minutes, while the value for FY 2005 was 642 minutes. Results from Phase 1 indicated that the weighed average length is slightly greater than the unweighted average length. Therefore, the rule-of-thumb estimate of 11 hours duration (660 minutes) for the average audio book title is still a very good value to use for planning purposes.
- Total Narration Cost: was $5,076,472 for FY 2005, and has averaged almost exactly $5 million per year for the past 5 years. A few data records in the PICS extract lacked narration costs, as cited in the footnote in Appendix 1, but very few.
- Average Narration Cost per Title: was $2,970 during FY 2005, and has averaged $2,841 for the past 5 years; NLS has noted that this cost is slowing increasing in recent years. These estimates are based upon calculations which excluded records with missing narration costs, as cited in the footnote in Appendix 1.
- Total Duplication Cost: was $3,440,136 for FY 2005, and has averaged $3,910,195 for the past 5 years; during FY 2005 slightly fewer titles were produced than usual, at slightly fewer copies per title than usual, and unit duplication cost has declined slightly simultaneously. This total excludes the cost of RC containers, which are provided as Government Furnished Equipment (GFE) to RC book producers, and cost on average about $0.55 each when new; this price is now approaching $0.60. About 1.4 million of the approximate 1.85 million containers required for mass-duplicated RC books annually are purchased, while the remainder of the requirement is provided from reconditioned containers (which cost $0.33 each to recondition, or $0.06 to inspect and dispose). About $950,000 per year on average has been spent over the last several years on both the purchase of new containers and the reconditioning and inspection/disposal of used containers. This cost will probably increase by about $100,000 for FY 2006 because of the higher unit price for new containers.
- Average Duplication Cost per Title: was $1,899 during FY 2005, and has averaged $1,991 for the past 5 years. This value excludes container costs, which added about $501 per title ($0.55 per container for 911 copies per title) in recent years. The unit price for a new container is expected to rise to $0.60 within the year.
- Average Duplication Cost per Copy: was $2.05 during FY 2005, and has averaged $2.16 for the past 5 years. This value excludes container costs, which added about $0.55 per copy during the time period. As noted, the new container unit price is approaching $0.60.
2.1.2 Audio Book Circulation
RC book circulation in the national library program has averaged approximately 20,000,000 copies per year over the last five years, although the values for both FY 2004 and FY 2005 were approximately 19,000,000 copies. Either value is a good estimate to use for long-term planning, even though possibly 5% of audio book circulation will continue (though not indefinitely) to be on RC for non-converted titles even after steady-state DTB operations have been achieved.
Appendix 2 shows RC circulation statistics reported to NLS by network libraries for FY 2005, which include activity for both RLs (net of associated SRL activity) and SRLs. The table shows annual and daily (assuming 250 working days per year) RC circulation data for individual readers (about 95% of total circulation), institutional readers (the institution itself being a patron), interlibrary loan circulation, and total circulation.
The libraries in Appendix 2 are listed in descending order of total circulation, and the table shows the relative ranking of each library. Both Multistate Centers (MSCs) have been excluded from the ranking, and their Inter-Library Loan (ILL) circulation prorated among libraries on the basis of their total circulation, but music and overseas audio circulation are included in the ranking. Network libraries to which site visits were performed are highlighted in bold.
Activity ranges from a high of almost 1,100,000 in Southern California (CA2A), to a low of 400 for the Virgin Islands (note: the next lowest is 5,300, with 10 libraries falling below 10,000 copies per year). Summary statistics are shown at the bottom of the table; the average (unweighted) annual circulation is about 140,000, the median is about 56,000, and there is great variability. The weighted-average annual circulation is almost exactly the same as the unweighted average. Wyoming is not included in the table since RC circulation services are provided entirely by another state. The data in Appendix 2 make clear the profile of library audio book circulation, which is not normally-distributed but rather has a large number of small operations, a moderate number of larger "mid-size" operations, and a small number of large operations.
2.1.3 Audio Book Collections
Appendix 3 shows RC collection size (i.e., number of copies) statistics reported to NLS by network libraries for FY 2005. The libraries are listed in descending order of total RC collection size, and the table shows the relative ranking of each library. Both MSCs and NLS-distributed music and books to overseas patrons have been excluded from this ranking. Network libraries to which site visits were performed during the project are highlighted in bold. Adjustments were made to FY 2005 reported values to compensate for apparent data capture errors for three libraries, and FY 2004 values were used as estimates for two libraries which did not report collection size for FY 2005 (but had RC circulation), as cited in the footnotes in Appendix 3.
RC collection sizes range from a high of over 2,300,000 copies for the Florida RL (FL1A) to a low of about 4,400 copies for the Gainesville, GA (GA1H) SRL. Summary statistics are shown at the bottom of the table; the average RC collection size is about 149,000 copies, while the median is about 48,000, and there is great variability. As was the case for circulation statistics, Wyoming is not included in the table since it does not have an RC collection and circulation services are provided by another state. The data in Appendix 3 show essentially the same type of profile for RC collections as that for RC circulation, i.e., a large number of small collections, a moderate number of larger "mid-size" collections, and a small number of large collections.
2.1.4 Audio Book Collection Turnover
Appendix 4 shows RC collection turnover, i.e., the circulation-to-collection ratio, for network libraries, derived from circulation and collection statistics reported to NLS by network libraries for FY 2005. The libraries are listed in descending order of collection turnover, and the table shows the relative ranking of each library. Both MSCs and NLS-distributed music and books to overseas patrons have been excluded from this ranking. Network libraries to which site visits were performed during the project are highlighted in bold. Adjustments for apparent data-capture errors for three libraries, and the use of FY 2004 values for two libraries, previously mentioned, are cited in the table footnotes. Two libraries that did not report collection size for FY 2005 (or for FY 2004) are excluded from the ranking.
RC collection turnover is a function of both circulation activity and collection size. Circulation activity is influenced by such factors as the size of the readership base, profile-select applications, Reader Advisor (RA) service levels, etc. Collection size is influenced by the original copy allotment quantities that libraries receive of new titles, attrition of copies through loss and damage, and weeding. Thus a number of factors determine the value of this inventory turnover ratio.
RC collection annual turnover ranges from a high of 4.2 for the Carterville, IL SRL (IL1B) to a low of 0.2 for the Detroit, MI SRL (MI2B). Summary statistics are shown at the bottom of the table; the average (unweighted) RC collection annual turnover is 1.3, while the median is 1.1, and there is variability; the weighted average is 1.0. Wyoming is not included for reasons previously noted.
The median and weighted-average RC collection annual turnover ratios, which are 1.1 and 1.0, respectively, are indicative of a very mature collection with an excess number of copies for many titles in many libraries. This conclusion is based upon the average turnaround time of circulated RC copies of 35 days, which translates into 10.4 inventory turnovers per year for continuously circulating copies. While the 10.4 turns per year is a theoretical maximum, only applies to the most popular bestsellers in their first year or possibly two of circulations, and is not a realistic target for average collection turnaround, the gap between the 1.0/1.1 actual system wide averages and this maximum indicate the potential for additional weeding of copies of titles of low demand beyond the levels currently employed.
Appendix 5 shows RC collection turnover by type of network library, i.e., turnover for RLs and SRLs separately. Given that many SRLs maintain RC collections for only the most recent years of production (and hence have smaller collections while circulating the titles that are relatively the most popular), while RLs maintain collections that span the full breadth of the collection (and hence have larger collections while circulating titles of all relative popularity), this comparison was necessary. The weighted-average annual collection turnover for SRLs is 1.5 versus 0.9 for RLs; unweighted averages are virtually identical for each library type, i.e., 1.5 and 1.0, respectively. The highest turnover for any RL is 2.3, while the highest value for a SRL is 4.2. The lowest values for SRLs and RLs, respectively, are 0.2 and 0.3.
2.1.5 Audio Book Collection Turnover by Title Age
In order to examine the turnover of the RC collection by title age, a special analysis was performed using the detailed title-specific circulation and collection data provided by Keystone Systems for use in Phase 1 of the project, the results of which are shown in Appendix 6. The analysis was done using FY 2004 RC circulation data and RC collection data current as of the date of the data extract, i.e., early in FY 2005. Data for eight RLs ("Ohio" consists of both Ohio RLs) and one SRL (NY) were used. Date Shipped by title (which is how title age was determined) was obtained from the PICS system.
As previously mentioned, both circulation activity and collection size influence the value of this ratio. Circulation activity generally declines strongly with title age, other things being equal and as noted in the Phase 1 report, as the relative popularity and hence demand wanes. Library policy/procedures can also influence this ratio, e.g., by constraining Profile Select-generated circulation to newer titles—which some libraries do. One or, if applicable, both factors result in declining circulation with title age. But collection size also declines with title age due to loss, damage, and weeding.
Two things are apparent from review of Appendix 6. The first are differences among the libraries with regard to weeding and, possibly, circulation strategies (e.g., constraining or leaving unconstrained profile select-generated circulation); since all libraries have the same title offerings (and obtain them at the same times), and all are subject to roughly equal loss and damage attrition, the differences arise due to different weeding and/or circulation strategies. The second is that, overall, there is potential for more weeding than is taking place on a system-wide basis. Generally, those collection age ranges with annual collection turnover values below 1.0 possibly have more copies than needed to support operations at a high service level; those falling below 0.5 certainly have a considerable number of excess copies.
2.1.6 Audio Book Collection Size by Title Age
In order to examine the size of RC collections by title age and how that varies among libraries, another special analysis was performed using the title-specific circulation and collection data provided by Keystone Systems cited above. The results of this analysis are shown in Appendix 7.
The analysis used RC collection data current as of the date of the data extract, i.e., early in FY 2005. As noted previously, data for eight RLs and one SRL were used, and date shipped/title age was determined from PICS system data. The original size of copy allotment for each age class by library was estimated based upon each library’s total copy allotment for FY 2004 (from the Keystone Systems data extract) adjusted by annual system wide production levels of RC copies for FY 2004 and for each year in the time period (from the PICS data extract). This approach was used because the actual original copy allotment quantities by library and by year were unknown.
As mentioned, collection size declines with title age due to both loss and damage attrition (which is uncontrollable), and to weeding (which is controllable). Two things are apparent from review of Appendix 7. The first are differences among the libraries with regard to weeding procedures, and possibly, circulation strategies (e.g., if Profile Select spans the entire collection in terms of age, then relatively more copies of older titles must be maintained in the collection versus that of a library which constrains circulation to some subset (in age) of the collection). Since all libraries are subject to roughly equal loss and damage attrition rates, the differences arise from different weeding and possibly different circulation strategies.
The second thing that is apparent is the potential for more weeding than is taking place on a system-wide basis. This conclusion is reached by comparing values for the "Combined" profile in Appendix 7 with the Demand by Age profile shown in Appendix 39 of the Phase 1 final report. The proportion of demand generated by the older age classes of titles falls to a much lower percentage of initial demand than does the ratio of collection size to original copy allotment size. In general weeding, while conducted, doesn't proportionately keep pace with waning demand because it requires incremental labor in addition to that required to perform (most) libraries’ routine book distribution tasks.
2.1.7 Library Information Systems
Appendix 8 contains a table with a listing of all network libraries including name, unique NLS Library Code, library type (RL or SRL), and type of information system used during FY 2005 (or early FY 2006), compiled from documents provided by NLS. The records are sorted in ascending order by Library Code, and libraries to which site visits were performed are highlighted in bold type.The information in Appendix 8 was then merged with FY 2005 RC circulation and collections data compiled by NLS, and summary results were derived for the different information systems, which are shown in Appendix 9. Six types of information systems were in use during FY 2005 for the 136 entities (132 libraries, 2 MSCs, and NLS overseas and music operations), and four libraries used no information system. The Keystone Library Automation System (KLAS) and Reader Enrollment and Delivery System (READS) have, by far, the greatest number of user sites (61 and 51, respectively), with the Consortium of User Library (CUL) system having 14 sites. The Albany and Austin systems are unique, or "independent," systems. It is anticipated that the SIRSI system is "going away" in the near future, two of the remaining four libraries using it have changed to other systems during FY 2006, and two more are moving towards implementation of a new system.
In terms of the proportion of RC circulation and collections supported/managed by the different systems, the KLAS libraries account for almost two-thirds (65.2% of circulation and 66.4% of collections) of the totals. After KLAS, the CUL libraries account for 12.5% of circulation and 14.1% of collections, followed by READS libraries with 12.3% of circulation and 10.6% of collections. After these, the Austin, Albany, and four combined SIRSI libraries account for 4.3%, 3.0%, and 2.4% of total circulation, respectively, with collections being in approximately the same proportions. Libraries with no information systems account for 0.3% of circulation and 0.6% of collections. Thus it is apparent from comparison of the proportion of sites to the proportions of circulation and collections that: the KLAS system is being used by relatively larger operations; the READS system by smaller operations; the CUL and SIRSI systems by mid-size operations; and the two independent systems by large operations.
Appendix 10 contains a table which lists several general characteristics of the five library information systems expected to be used in the future system (SIRSI is not included) as provided by the respective information system vendors and representatives in a meeting with ManTech held at NLS. The characteristics included are: Internet Access (Outgoing); Operating System; Database; Programming Staff to make changes necessary for implementation in a future system; and Remote Internet Accessibility (Incoming).
- Internet Access (Outgoing): With the exception of a few READS sites, all network libraries that have information systems have an outgoing Internet connection with "T1" speed or greater. RE the few exception READS sites, a few have modem connections to the Internet, while a very few sites have no connections.
- Operating System: the CUL sites use Windows 2003 Server – XP Server, Visual Basic and ASP; the Albany system uses Solaris 9 with Visual C++; the KLAS sites use Windows 98 - XP, SUN Solaris v.8-v.10, and Linux Red Hot; the Texas system uses Debian Linux; and the READS libraries use Windows 98 and XP.
- Database: the CUL sites use SQL Server 2000/ODBC; Albany uses Oracle 9/ODBC; the KLAS sites use Progress/ODBC; Texas uses Unify (but will be changing within a year); and READS sites use MS SQL Server 2000/ODBC and MS Access.
- Programmer-On-Staff: Only Texas and Albany have Programmer-Analysts on-staff to modify their information systems. Changes to the KLAS system are performed by Keystone Systems, changes to the CUL system are performed by CUL system consultants, and changes to the READS system are performed by Data Management Associates. For all five information systems, all Programmer-Analysts who make changes to the systems have experience working with XML.
- Remote Internet Accessible (Incoming): Most CUL libraries can be accessed remotely, but a few cannot. Neither Albany nor the Texas system can be remotely accessed by the Internet; the Albany system has a firewall and access can be enabled; the Texas system, with a firewall, can enable access; the KLAS system can be remotely accessed; and READS has remote access, with firewall.
2.2 Site Visits to Network Libraries
Site visits were performed at a sample of network libraries in order to gain a thorough understanding of audio book distribution operations in the network, including the variations therein. Appendix 11 lists the site visits made during Phase 2 of the project (the Maryland RL, using the SIRSI system, was visited during Phase 1).
The sites were selected to provide examples of the functioning of all five information systems expected to be in use in the future system and, while focusing on RL operations, also observe at least two SRL operations. Of particular interest, aside from representation of a specific information systems, were: (1) the Albany, NY operation, in which many audio book distribution tasks are automated; (2) the Austin, TX operation, in which the entire audio book collection is stored using a random storage method; and (3) the New York, NY operation, in which the book collection is stored using a terminal-digit storage method. The Lansing, MI RL was selected because it is the only CUL library that uses SRLs. The Daytona Beach and Jacksonville, FL sites were selected as representative KLAS libraries which operate within an integrated distribution system. The Charleston, WV and Fairfax, VA libraries were selected as the representative READS RL and SRL, respectively.
2.2.1 Albany, New York Regional Library
These observations and discussions were made while in the company of Jane Somers, the Regional Librarian; Van Judd, the Director-Program Manager; and David Gosda, the IT Director. The purpose of our visit was to witness and critique the mechanized Virtual Turnaround System (VTS) used for processing and shipping customer returns, and to familiarize ourselves with the proprietary LAS that is used. We also discussed a formative transition plan for integrating DTB distribution into library operations with all parties.
1. USPS Interface
The New York Regional Library in Albany is located in the Museum/Archives/Educational building of the Campus Complex, and the building has its own zip code. Docking facilities are shared, and the library receives two drop off/pickups per day, one at 10:00 AM and the second at 2:00 PM. The Library has a good rapport with the USPS.
All RC input is received from the Albany Bulk Mail Center (BMC), and all RC output is sent to the Albany BMC. The handling unit for this workflow is a large metal open-top tow-line container called a BMC (Bulk Mail Container, also referred to as an “OTR”). A BMC measures 42" wide x 60" long x 69" high, and will hold 1,100 RC containers. About 20 BMCs are received and processed each week, or about four per day. The incoming BMC’s also contain new receipts, braille, machines, patron non-deliveries, and other packages, which must first be sorted out.
2. Bulk Container Unloading
The BMC’s are unloaded using a specially designed dumping machine that does not require a high lift. The machine lifts the BMC some 3" off the floor and then rotates it slowly to dump the contents onto a 36" wide heavy-duty pan-type belt conveyor that rises to table height to push the books onto a slick-steel sort table. The table is 10 feet long and 36" wide, and has 3" high side rails on 3 sides. The conveyor belt and tilt mechanism on the dumper are manually controlled to keep the work table full. Packages other than returned RC’s are then removed. The new RC’s are taken to a separate area for receiving, and the returned RC’s are fed to the VTS on a nearby conveyor.
3. Operating Schedule
The library operates from 8:00 AM to 4:30 PM. There are three scheduled 30 minute work breaks - one in the morning, one at noon and one in the afternoon. The pick tickets for order filling from collection storage are printed at 7:30 - 8:00 AM. The picking is done all day long and requires an average of 2 FTE. Patron orders that cannot be sourced to collection storage are backordered in the VTS.
4. Daily Throughput
The library receives and ships 2,250 RC copies per day, on average. Some 50% of the shipments are from collection storage and 50% are from the VTS. The VTS throughput, which includes the re-circulation of receipts, is “3,000 - 4,000 RCs per day, 5,000 maximum.”
5. RC Collection Size
The library has some 28 years of RC books, ranging from RC 06000 to RC 61000. There were 617,000 RC copies in the collection as of March 14, 2006, and only 26.7% of the copies were actually in circulation. Some 55% of the patron “Has Now” records are over one year old.
6. In-House Duplication
The library has high speed cassette duplication capabilities, and has a full collection of RC masters, which is located on another floor. Patron returns that are damaged are terminated at VTS inspection and some copy replacement is required. A typical replacement order is 5 - 7 copies. Open orders for titles that are back-ordered are shipped to patrons directly from duplication.
7. Container Bar Coding
The containers are bar coded, and the bar coding is used to track RC copies to and from patrons and to manage inventory.
8. Library Automation System
The library uses its own proprietary Library Automation System.
9. Pick Ticket/Address Cards
The Pick Ticket/Address cards for patron orders that are filled from collection storage are charged to the patron Has-Now record by the system; and the pick tickets are printed in stock location (title number) sequence. The pick ticket also contains a bar-coded transaction number that provides a bridge to the patron number and the copy number. This bridge is made when the transaction number barcode on the card and the container barcode are scanned.
The pick ticket/address cards for patron orders that are filled by the VTS contain only the patron name and address, as container movement to the patron is automatically tracked by the VTS system via the container barcode. The VTS also receives all patron returns by scanning the bar codes on the containers. The patron record Has-Now and Has-Had status are changed, and the title inventory record is charged.
10. New Book Receiving
The receiving clerk checks and updates the system screen for each new title as part of the receiving process. Bar-coded labels are then affixed to all but the master copy, and the bar codes are scanned into the system. The copies are then placed into custom-cut corrugated boxes that hold 16 copies. These boxes are sized to fit in the 7-wide collection shelf openings. However, the new receipts are taken not to collection storage, but are placed in a hold area pending release of the titles for circulation. The new titles are later picked and shipped to patrons from the hold area.
11. Collection Storage
Container storage in the stacks is by title number, in library-type shelving. Most storage is 2-deep, but there is a sizable amount of 1-deep shelving. The library plans to use some of the 1-deep shelving for DTB storage.
12. VTS Receiving
The VTS is rated to process 12 copies of returned RC’s per minute for a full 8 hours a day. However, it operates at lower capacity much of the time, but processes all required throughput. The bottleneck in the system is the inspection/rewind operation, which is now being studied for improvement; it is manual-intensive, and extremely difficult to automate.
Returned RC’s enter the VTS from the receiving area (sorting table) on a narrow belt conveyor and the containers are oriented on the conveyor with the straps forward and up, and the address card holder down. The first workstation on the conveyor is container opening. A container is first singulated into the station with an electric-eye that brings the container forward and holds it firmly in position. A mechanical arm with a claw-shaped end then proceeds to unsnap the two buckles on the container, but the lid of the container is not opened. The container then moves out of the opener station and onto a carousel conveyor that feeds the inspection/rewind workstations.
There are four inspection workstations on each side of the carousel conveyor, and the containers continue to circulate on the conveyor until removed by the inspectors. RC copies that are found to be damaged are set aside for termination, rather than being reconditioned. Copies that pass inspection are placed on a takeaway conveyor and proceed to the next workstation, which is fully mechanized.
This work station performs three functions in sequence. It first closes the straps on the container, then removes the address card on the container, using a vacuum extracting arm. Finally the container is turned upside-down so that the address card holder on the container is up. The container then moves onto a takeaway conveyor, which accumulates and singulates the containers to provide an even workflow for the downstream operations. Here the input operations end and the output operations begin.
13. VTS Shipping
The remaining operations on the VTS are performed while the containers are in motion, and the processing speed is 12 copies per minute. The first operation is to read the barcode on the container to determine if there is an open order for the title (that can thus be fulfilled on the VTS), and to change the patron record for the return from Has-Now to Has-Had. The required response time for this reading is one second. If there is an order for the book, a blank address card is automatically placed into the card holder on the container. The final operation is to print the patron’s name and address on the card that has been placed on the container. Two high-speed inkjet print heads are used to perform this operation, but the printers do not print the postal bar code of the patron.
The containers for filled orders continue on the conveyor until they are deposited in a BMC for shipment, and containers that have no orders are shunted off for recirculation through the VTS. These containers are held for reprocessing for a minimum of two days, and are reentered into the system prior to the barcode reading station (having already passed inspection, with cover closed, straps fastened, and old card removed). After the second pass through the system, the containers with no orders are automatically sorted into bins by the first digit of the title number. After screening, these copies are either returned to collection storage or terminated. Most of these cleanup operations are done on Friday.
2.2.2 Austin, Texas Regional Library
These observations and discussions were made while in the company of Edward Earle, the Warehouse Manager, who escorted us through the facility. The purpose of our visit was to witness and critique the random storage mode used for the RC collection, and to familiarize ourselves with the proprietary LAS that is used. We also discussed a formative transition plan for integrating DTB distribution into library operations with Edward Earle; Ava Smith, the Regional Librarian; and Robert Helfer, the Automated Systems Administrator.
1. Building Configuration
The Texas Distribution Center (DC) is in a 1-story warehouse-type building that has a large storage mezzanine. The DC is located three miles from other library operations, which are in the Capital Complex in downtown Austin. Included are data processing, the Reader Advisors, management offices, in-house duplication, title master storage, and publishing.
2. USPS Interface
Texas has no subregional libraries, and substantially all interface is with the BMC in Dallas. Although all incoming receipts come from the Dallas BMC, they are routed through the Sectional Center in Austin. The USPS delivery truck arrives about 10:30 AM for delivery/pickup, and the returning RC’s and new books are in cage carts. A cage cart is a 3-sided cart that measures 42" wide x 29" deep x 70" high and holds 700 RC containers. The fourth side of the cart is made of heavy canvas, which is partially removable to facilitate loading and unloading. The loads to Dallas are also in cage carts, but books destined for the Austin and San Antonio areas are separately loaded in hampers.
3. Operating Schedule
The warehouse works from 8:00 AM to 4:00 PM. The LAS runs an overnight batch routine to fulfill reserves, requests and profile-select orders; the routine typically takes 1.5 hours, with a maximum of 2 hours. Printing of pick tickets occurs first-thing in the morning, and takes over an hour, but less than 2 hours. Picking of orders begins at 8 AM and is completed by 3:30 PM. Processing of new RC titles, and inspection and processing of returned RC copies, goes on until operations end in the late afternoon. All customer returns must be processed and available for shipment to patrons within 48 hours.
4. Daily Throughput
An estimated 3,200 copies are received and shipped in an average day. Some 85%-90% of the copies shipped come from the turnaround area, referred to as Area or Zone "A."
5. RC Collection Size
The warehouse carries 25 years of title production and about 569,000 RC copies. Container storage is strictly random by copy number in both the turnaround and collection storage areas. The containers are stored 1-deep in 11" deep shelving. They were stored 2-deep prior to a major restructuring, which is nearing completion. This transition was made possible by better inventory management.
6. In-House Duplication
Most of the in-house duplication is for Texas publications, especially magazines. However, the Reader Advisors can order a copy of a book at any time. The duplication process is also system-assisted, in that the system prioritizes in-house duplication needs. Patron returns that are damaged are evaluated, and then either terminated in entirety or the container is salvaged and used with replacement cassettes if the title is in high demand.
7. Container Bar Coding
The containers are bar-coded. The barcodes are used to control the storage and movement of every copy in the warehouse, as required for random storage. There are over 700,000 copy location records in the database.
8. Pick Ticket/Address Cards
The 3" x 5" pick tickets/address cards contain a bar-coded transaction number that is assigned by the computer during the order filling run, and the tickets are printed in stock location sequence. The transaction number provides a bridge to the title number and the patron number. It is used to track a book from the time of shipment to the time of return. The Has-Now patron record is charged and the inventory decremented when the transaction barcode on the pick ticket and the barcode on the container are scanned, and the Has-Had patron record is charged and the inventory is incremented when the barcode on the container is scanned.
9. New Book Receiving
New books are received at a workstation near the receiving dock. Here the containers are bar-coded and entered into the system and a return address decal is placed on every container. The books are then taken to collection storage and placed on location (which is Area or Zone "N"). This location entry makes the books available for distribution.
10. Returns Receiving
Customer returns are placed in Quickturnaround (Zone or Area "A") storage after having passed inspection, in Area "H" (Hold) if awaiting inspection, or moved to Area "Z" after inspection if damaged. In any event, the barcode on the container is scanned and the barcode of the shelf location is scanned, which simultaneously relieves the patron record and increments inventory. This data entry changes the patron record from “Has-Now” to “Has-Had”, but only makes the copy available for circulation if in Area A (if not inspected or damaged, it is not available for circulation).
11. Returns Inspection
Returned RC’s are sometimes temporarily stored in collection storage awaiting inspection (Area H). The copies are not placed on location as they have not been officially received. The inspection is done by volunteers, rather than the warehouse staff, and all returned books are inspected within 48 hours of receipt. RC’s that are found to be damaged are usually terminated rather than reconditioned.
12. Turnaround Shelving Storage
The turnaround books are stored in 18 shelf ranges. A range consists of 26 sections of back-to-back shelving 36" wide x 11" deep, and the aisles between ranges are 36" wide. There are 5 shelf openings in a section and each shelf opening will hold five stacks of 8-high RC’s. Each stack is a separate location in the random storage locator system. Austin plans to use the same configuration for DTB’s, but will install more shelves so that the present 8-high container stacking is maintained. To place a copy on location, the barcode on the container and the barcode of the shelf are scanned. The scanner that is used for these location entries has a large memory and the entries are later batch transmitted to the system in downtown Austin.
13. Turnaround Stock Rotation
The working inventory of turnaround stock is about a 32-day supply. When more storage space is needed for incoming receipts, the warehouse manager will use a computer program to find the range having the lowest inventory. The stock in this range is then weeded to remove Texas publications and RC’s with the lowest title numbers, which copies are then relocated to slower moving storage zones. The containers remaining in the range are then consolidated, or "compressed," and their new locations are entered into the inventory records.
14. RC Collection Storage
The library stores about 456,000 RC copies at any given time. Warehouse management staff initiate the movement of excess copies from faster moving to slower-moving storage zones.
15. Weeding of the Collection
Weeding in the operation is system-assisted, and uses both copies in collection and circulation activity to determine which titles to weed and how many copies should be removed. Weeding is performed continuously, and the system generates weeding pick tickets. The XESS protocol is utilized, as are disposal shipment to York.
2.2.3 New York, NY Regional Library
These observations and discussions were made while in the company of Bonnie Farrier, Principal Librarian, who escorted us through the facility. The purpose of our visit was to witness and critique the terminal digit storage mode used for the RC collection, and to observe the KLAS system in operation, particularly, the operating procedures followed in the turnaround area. We also discussed the formative transition plan for integrating the DTB product line into ongoing library operations, with both Bonnie Farrier, and the Regional Librarian, Robert McBrien.
1. Building Configuration
The library is in a 5-story building, with only the receiving/shipping dock and a returned book inspection area on the first floor. Offices are on the second floor; receiving and turnaround shelves are on the third floor; reproduction and master copy storage are on the fourth floor; and the terminal digit storage collection is on the fifth floor.
2. USPS Interface
The USPS delivery truck arrives at 9:00 A.M. Shipping pickup is done by the New York City Library truck, which arrives at 5:00 A.M., so as to avoid the morning traffic congestion. The RC’s are received and shipped in cage carts.
3. Operating Schedule
The library operates on a 1-shift schedule, from 8:00 A.M. to 5:00 P.M. Container receiving, inspection and putaway in the turnaround shelving are done in the morning and early afternoon and order filling and shipping are done in the late afternoon. The computer order-filling run for the next day’s shipments is done in the evening, and the pick tickets are printed and burst by a contractor.
4. Daily Throughput
An estimated 1,000 copies are received and shipped on an average day. Some 60% to 80% of the copies shipped were said to come from the turnaround area.
5. RC Collection Size
The library currently carries only 5 years of title production, but title masters are retained for 25 years. Container storage is strictly by year and by title number, in 2-deep shelving. (See the later discussion of terminal digit storage for more details).
6. In-House Duplication
Most of the in-house duplication is for New York State publications. NLS titles are kept on Profile-Select for only 5 years and the only demand for older titles is therefore by direct patron request. This is some 5 to 10 copies per day, which are duplicated-on-demand.
7. Container Bar Coding
The containers are not barcoded.
8. Library Automation System
The library uses KLAS, Version 5.
9. Pick Tickets/Address Cards
The 3" x 5" pick tickets/address cards contain a barcoded transaction number that is assigned by the computer during the order filling run. The number provides a bridge to the title number and the patron number. It is used to track a book from the time of shipment to the time of return. The "Has-Now" patron record is charged when the pick ticket is printed, and the "Has-Had" patron record is charged when the address card on a returned book is scanned.
10. New Book Receiving
New books are received on tables in a separate area on the third floor. After data entry, the books are usually taken to collection storage on the fifth floor, and are placed in a hold area, pending release of the titles for circulation. The titles are then picked and shipped to patrons from the hold area.
11. Returns Inspection
Returned books are first inspected in the receiving area on the first floor, at which time, all books with irregularities are set aside for termination. All other copies are then sent to the third floor, where they are rewound by the receivers.
12. Returns Receiving
Customer returns are received on tables in a separate area on the third floor, and the process is usually completed by noon. The receiving consists of scanning the barcode on the address card. This data entry changes the patron record from "Has-Now" to "Has-Had," and charges the inventory record. The containers are then randomly placed in numbered tote trays that are sized to fit on the shelves of the carousel conveyor that is used for turnaround shelving. There are 16 4-cassette containers in a tray and the address cards for the books in each tray are banded and placed in the tote for reference, if necessary. The pick tickets for the turnaround shelves are then later printed in the same sequence in which the numbered tote trays are stored.
13. Carousel Turnaround Shelving
There are four carousels, but only one is needed for the RC turnaround shelving. The library is planning to use the other three carousels for DTB storage. A carousel has 20 carriers, and a carrier has 12 shelves, with the totes stored 3-wide. The bottom shelf on the carrier is not readily accessible and is not used. The maximum capacity of each carrier is therefore 528 4-cassette containers and the maximum capacity of a carousel is 10,560 containers. Only a 1-day supply of books is kept in the turnaround shelving, as any copies that remain after order filling are taken immediately to collection storage, where they are sorted by the last two digits of the title number, and shelved.
14. Terminal Digit Collection Storage
As noted, there are 100 2-digit storage modules, and each module has six shelf sections. Each shelf section is six shelves high and each shelf opening has seven 8-high container stacks, stored 2-deep. The maximum storage capacity of a shelf section is therefore 672 containers. The maximum storage capacity of each 2-digit storage module is 4,032 containers.
Most new titles are now ordered from the duplicator in lot sizes of 16 or 32 copies, but only one storage slot is assigned to a title. As demand wanes, the storage slot will become full, and the surplus copies are then terminated. This is a simple and highly-effective form of weeding.
Space is provided for new books by weeding and consolidation of the individual storage modules, as needed, such that the space allocated for collection storage is fully utilized. This selective weeding and consolidation will not disrupt the general workflow, but all 100 storage modules will have to be weeded and consolidated as often as necessary.
15. RC Collection Storage Capacity
The New York Library has recently completed a major restructuring, in which the inventory of all title numbers from 10,000 to 50,000 was terminated. The present storage modules, as described, now contain five years of titles and are about half full. There should therefore be sufficient storage capacity to accommodate all new RC books through circa 2010, without any adjustment of inventory levels.
2.2.4 Lansing, MI Regional Library
These observations and discussions were made while in the company of Sue Chinault, Librarian, and her staff, who escorted us through the facility. The purpose of our visit was to witness the CUL system in action, particularly the turnaround methodology and the RL/SRL relationships. We also discussed the formative transition plan for integrating the DTB product line into ongoing library operations with all parties.
1. Building Configuration
The library is located on the first floor of the State Museum building, and occupies some 26,000 square feet. The receiving and shipping docks are also on the first floor, but are shared with other building occupants.
2. USPS Interface
The Michigan Department of Management and Business truck delivers books to the local post office and picks up books for return to the library. Receiving is at 9:00 - 9:30 AM and shipping is at 12:30 PM.
3. Operating Schedule
The library operates on a 1-shift schedule from 8:30 AM to 6:00PM. Order picking and shipping are necessarily done in the morning to meet the 12:30 PM shipping deadline. Receiving, inspection, check-in and putaway begins at 9:00 AM and continues throughout the day. The overnight pick ticket order filling run begins at 9:00 - 10:00 PM and takes about 5 hours. The pick tickets take 1 - 1.5 hours to print, and separate runs are made for the turnaround area and collection storage.
4. Daily Throughput
Throughput averages 800 RC copies per day, and this total includes shipments made on behalf of the 10 subregional libraries.
5. RC Collection Size
The library has a 25-year inventory of books and the collection size is 232,000 copies.
6. In-House Duplication
The library prints only one local magazine. Replacement cassettes are reproduced and relabeled, as needed. There is no facility to produce container labels, but they are not often needed.
7. Container Bar Coding
The containers are not bar coded.
8. Library Information System
The library and 8 subregional libraries use CUL. The 2 other SRLs now use READS, but are being converted to CUL. The RL/SRL’s use a common server, and the libraries are "Internet enabled", e.g. staff at the RL can operate an SRL printer.
9. Pick Tickets/Address Cards
The 3" x 5" pick tickets/address cards contain a bar-coded transaction number that is assigned by the computer during the order filling run. The number provides a bridge to the title number and the patron number. It is used to track a book from the time of shipment to the time of return. The "Has-Now" patron record is charged when the pick ticket is printed, and the "Has-Had" patron record is charged when the address card on a returned book is scanned.
10. New Book Receiving
New books are received in a separate area, and will be held in the area until all copies of a title have been received and accounted for. The copies are then placed on location in the stacks.
11. Returns Inspection
Returned books were once inspected by students, but are now inspected by the staff. This inspection is done before the books are placed on the tables that are used as a turnaround area, and the copies needing repair are set aside. The rewinding is done later by the receivers.
12. Returns Receiving
The inspected books are placed in stacks of 16 on the turnaround tables to begin the formal receiving. The address cards are then removed from all the containers in a stack and banded together. The cards are then assigned a serial stack number, following which the stack is slid onto a second table which indicates that the stack has been accounted for. The banded cards in all stacks are kept in stack number sequence, and the cards are later scanned in that order to indicate receipt. This scanning changes the patron record from "Has-Now" to "Had-Had", and charges the inventory record. The pick tickets for the turnaround area will later be printed in the order in which they have been scanned, and any books that are left over in the turnaround area at the end of the day are considered by the system to be in collection storage.
13. Collection Storage
Collection storage is by title number in 1-deep, 6-high compact shelving and the lot sizes ordered range from 8 to 16 copies. Some shelving can apparently be made available for DTB’s, and we therefore demonstrated how the shelving can be reprofiled to 10-high to accommodate 43% more DTBs. In a Hybrid book distribution model, with the libraries being allocated and distributing only 800 fast-moving new titles per year, 14 shelf sections would be needed to store a full year’s inventory of DTBs.
The library prints only one local magazine. Replacement cassettes are reproduced and relabeled, as needed. There is no facility to produce container labels, but they are not often needed.
7. Container Bar Coding
The containers are not bar coded.
8. Library Information System
The library and 8 subregional libraries use CUL. The 2 other SRLs now use READS, but are being converted to CUL. The RL/SRL’s use a common server, and the libraries are "Internet enabled", e.g. staff at the RL can operate an SRL printer.
9. Pick Tickets/Address Cards
The 3" x 5" pick tickets/address cards contain a bar-coded transaction number that is assigned by the computer during the order filling run. The number provides a bridge to the title number and the patron number. It is used to track a book from the time of shipment to the time of return. The "Has-Now" patron record is charged when the pick ticket is printed, and the "Has-Had" patron record is charged when the address card on a returned book is scanned.
10. New Book Receiving
New books are received in a separate area, and will be held in the area until all copies of a title have been received and accounted for. The copies are then placed on location in the stacks.
11. Returns Inspection
Returned books were once inspected by students, but are now inspected by the staff. This inspection is done before the books are placed on the tables that are used as a turnaround area, and the copies needing repair are set aside. The rewinding is done later by the receivers.
12. Returns Receiving
The inspected books are placed in stacks of 16 on the turnaround tables to begin the formal receiving. The address cards are then removed from all the containers in a stack and banded together. The cards are then assigned a serial stack number, following which the stack is slid onto a second table which indicates that the stack has been accounted for. The banded cards in all stacks are kept in stack number sequence, and the cards are later scanned in that order to indicate receipt. This scanning changes the patron record from "Has-Now" to "Had-Had", and charges the inventory record. The pick tickets for the turnaround area will later be printed in the order in which they have been scanned, and any books that are left over in the turnaround area at the end of the day are considered by the system to be in collection storage.
13. Collection Storage
Collection storage is by title number in 1-deep, 6-high compact shelving
and the lot sizes ordered range from 8 to 16 copies. Some shelving can apparently
be made available for DTB’s, and we therefore demonstrated how the shelving
can be reprofiled to 10-high to accommodate 43% more DTBs. In a Hybrid book
distribution model, with the libraries being allocated and distributing only
800 fast-moving new titles per year, 14 shelf sections would be needed to
store a full year’s inventory of DTBs.
2.2.5 Charleston, WV Regional Library
These observations and discussions were made while in the company of Donna Calvert, Director – Special Services, and her staff who escorted us through the facility. The purpose of our visit was to witness the READS system in action, and particularly because Charleston exemplifies a regional library having subregional libraries that are not integrated. During our visit, we also discussed the formative NLS transition plan for integrating the DTB product line into ongoing library operations with all parties.
1. Building Configuration
The library is located on the ground floor of the State Archives Building, and has an area of 8,800 square feet. The receiving and shipping docks are also on the ground floor, and are shared with other building occupants.
2. USPS Interface
The USPS truck delivers at 8:00–8:30 AM and there are typically 2 hampers per day. Deliveries to the Post Office are made at 9:00 AM by library personnel using a van. The USPS Sectional Center is some 12 miles away.
3. Operating Schedule
The library operates on a 1-shift schedule from 8:00 AM to 5:00 PM. Receiving, inspection and check-in are done from 8:00 AM to noon, and putaway is done from noon to 3:00 PM. The batch order filling run is made at 2:45 PM and takes 10 – 15 minutes. Pick tickets are printed at 3:00 PM and takes about 5 minutes. Order filling is done at the end of the day and early next morning.
4. Daily Throughput
Throughput averages 310 RC copies per day, and this total includes shipments that are made on behalf of the four SRLs.
5. RC Collection Size
The library has a 25-year inventory of RCs, and the collection size is 85,000 copies.
6. In-House Duplication
Most of the in-house duplication is for 2 magazines. There is no master title collection, and patron orders for books that are not in stock are either forwarded to the MSCs, or a copy is ordered from the MSCs that serves as a master for making a circulating copy. RC duplication of NLS titles averages 4 copies a day.
7. Container Bar Coding
The containers are bar-coded and the first 3 digits of the barcode are library unique (which is the case for all libraries using READS). The barcodes are used for tracking books to and from the patrons.
8. Library Automation System
The RL and its SRLs use READS. Communication between the RL and its SRLs is by e-mail, i.e., the libraries do not share a common system server. However, all patrons in West Virginia are registered by Charleston.
9. Pick Ticket/Address Cards
The 3" x 5" pick ticket/address cards are copy-specific, and it is assumed that a copy is shipped when the ticket is printed. There is no bar-coded zip code on the ticket for either the patron or the library. The patron Has-Now record is changed and the inventory record is charged when the barcode on a returned container is scanned. There is a code on the pick ticket to indicate the specific SRL ordering the book, as appropriate. Profile-Select order filling is limited to titles RC45000 and up.
10. New Book Receiving
New books are received at a table in the receiving-shipping area at which time they are bar-coded on the side of the container and the barcode is scanned to charge the inventory record. The last 3 digits on the barcode (i.e., copy number) are marked on the label-end of the container, so as to be visible from the storage aisle. The books are then placed in storage.
11. Returns Inspection
Returns are inspected and rewound at workstations in the receiving-shipping area. This work is done by both staff and volunteers.
12. Returns Receiving
Returns receiving is done at tables in the receiving-shipping area, at which time the barcodes on the containers are scanned. The receiver also checks the address card to see if the book was shipped for a subregional library, and notifies the subregional library by email that the book has been returned. The returned books are then sorted by title number and placed back in storage.
13. Turnaround Storage
The library does not have a quick turnaround capability.
14. Collection Storage
Collection storage is by title number. There are 430 sections of 1-deep shelving and 40 sections of 2-deep shelving. The 2-deep shelving is used for the newest titles. There is generally only one copy per title for books under RC40000.
The library was recently told by the State Librarian to vacate part of their collection storage area, and as a result was required to terminate 23,000 copies, or about 25% of the copies in the collection. The library now has no room for new RC books or future DTB books unless RC inventory is further reduced or present space utilization is improved.
2.2.6 Shelf Storage Alternative for Small Lots of RC Containers
The proposed shelf storage mode considers that RC containers would be stored on their sides in title number sequence, rather than in shelf stacks, and is appropriate only for 1-deep container storage. This configuration is now used by some libraries to store their small lots. The term "small lots" refers to the lot sizes required for most of the older titles for all libraries, and for most of the titles in the smaller libraries. The storage mode and analysis technique suggested in Appendix 12 will automatically demonstrate what would be a small lot for every library, and quantify any potential increase in storage capacity.
If a decision is made to reprofile the present shelving, which will necessarily involve obtaining additional shelves, a few further suggestions are:
- Some libraries now store their master copies with the regular collection. These copies should be removed and stored in a separate area, but in the same manner as the regular collection.
- The implementation process will provide a timely opportunity to decide how many copies of each title should be retained and how many should be terminated.
2.2.7 Second Return Address Label on a Container
When the address card on a returning container is missing, the USPS will generally return the container to the nearest network library, which may not be the library of origin, and there are no doubt instances where the container is not returned at all. To remedy this problem, the Texas library affixes a permanent pressure-sensitive return address label to every container.
2.2.8 An Early Weeding Stratagem
Some libraries now assign only one storage slot to new books, regardless of the quantity ordered. The new books are placed in a hold area after receiving, from where they are shipped to patrons. The returning books are then intercepted in the turnaround area and recirculated to other patrons. This cycle continues until the number of copies returned exceeds the number of copies reshipped, at which time the remaining books are put into their assigned storage slots. When a storage slot becomes full, the surplus books are terminated. This is a simple, timely, and highly effective form of early weeding.