Cost Comparison

of
Automated versus Manual
Materials Handling Operations
at
King County Library System

Prepared
for
King County Library System
by
Lori Bowen Ayre
The Galecia Group
January 26, 2009

Table of Contents
Table of Contents ....................................................................... 2
Executive Summary ..................................................................... 3
Introduction .............................................................................. 4
Materials Handling Environment ................................................. 4
Workflow at Manual Library ...................................................... 5
Workflow at Automated Library ................................................. 5
Methodology .............................................................................. 7
The Problem with Using Performance Tests ................................... 7
Compare Similar Live Operations................................................ 8
Identifying Relevant Factors ..................................................... 8
Unavailable Resources ............................................................ 9
Frontline Staff Time .............................................................. 10
Backroom Staff Time ............................................................. 10
Customer Time .................................................................... 11
Results of Comparisons ................................................................ 12
Conclusion ............................................................................... 14

Executive Summary
In order to evaluate the benefits of automated materials handling (AMH)
systems, two similar libraries were compared (one with an AMH system and one
without). The study found that costs related to materials handling at the
automated library (library with an AMH system) were significantly lower than
the manual library (library with no automation).
Costs were compared in four areas: the cost of unavailable resources (items
such as books, DVDs, and CDs that are unavailable to customers because of
backlogs in processing), the costs of frontline staff time spent on materials
handling related requests, the costs of backroom staff assigned to materials
handling processing, and the costs to customers.
To determine the cost of unavailable resources, the author establishes a daily
use cost based on the average cost of acquiring a new library resource and the
lifetime of the item in the library collection. The study found that $12,950
more was spent per year on unavailable resources at the manual library.
The second cost area relates to frontline staff and the time spent dealing with
three recurring problems related to materials handling that were significantly
reduced at the automated library. The study found that $47,322 more was
spent on frontline staffing costs handling problems at the manual library.
The third cost comparison focuses on backroom staffing required to perform
materials handling functions. The manual library spends $75,712 more per year
than the automated library.
Finally, the study considered the cost of the extra time spent by a customer
going to the manual library expecting library material to be on the shelf
based on the library catalog. Though the material had been checked in, it was
waiting in a backroom unavailable to the customer. The study showed that
customers at manual libraries may spend as much as $84,845 per year on
unnecessary trips to the library.
The report establishes that automated materials handling results in significant
savings in four areas at KCLS. The report also notes that an automated
materials handling system would likely yield even greater benefits at libraries
lacking KCLS automated central sort operation which provides presorting and
tote check-in for interlibrary delivery.

Introduction
In October, 2008, King County Library System contracted with Lori Ayre to
assist them in evaluating the cost and benefits of their automated check-in
systems. Automated check-in systems are part of a larger automated materials
handling (AMH) system that also includes sorting of the material checked in.
The KCLS Board sought additional information about the costs associated with
the AMH systems versus the realized benefits. The goal was to engage a
consultant who would perform an objective study of the value of the AMH
systems being installed at KCLS branches.
At the time of study, KCLS had already installed AMH systems at six of their 43
libraries, and three more were in the process of being implemented. The
Library believed that the six systems had been in operation long enough that
any cost, time or staff savings associated with their use could now be
quantified.
Materials Handling Environment
To consider the effect of the AMH systems, the broader materials handling
environment at KCLS must be reviewed. This includes how interlibrary
deliveries are handled and how customer returns are handled.
Six of the KCLS libraries have AMH systems installed. A library AMH system is
composed of three components: a library sorter, the automated check-in
station(s), and a staff induction station. The automated check-in station allows
customers to induct their returns into the sorter which then checks in the item
and sorts it to one of several shelving carts or delivery totes. The induction
station functions like the automated check-in station but is designed for staff
use. Both feed into the sorter via a system of conveyors.
All KCLS libraries benefit from a large, central sorter used for interlibrary
deliveries. All sorting for interlibrary delivery is done in Preston. The Preston
sorter communicates with the integrated library system to determine the
status of each item and, based on the status, routes the item to the
appropriate tote. The totes are then delivered to each library daily.
The Preston sorter not only sorts all of the material to totes but also builds a
tote manifest which allows for tote level check-in at the library. When
libraries receive a tote via interlibrary delivery, they check-in all the items at
once (in a batch operation) by scanning just the bar code on the tote (instead
of having to scan all 40-50 items in each tote). Tote check-in ensures that
interlibrary delivery handling is very efficient.

Returns (items returned to each library by customers) are handled differently

from interlibrary delivery. With returns, each item must be checked in and
then rough sorted to a book cart for re-shelving. An automated library (library
with an AMH system) handles returns differently from a manual library (library
without an AMH system).
Workflow at Manual Library
At the manual libraries, most returns are dropped into book return slots. When
full, a bookdrop bin is moved to the check-in station and worked directly out of
it.
Check-in stations consist of a workstation with scanner (so each item can be
scanned into the system), printer (so hold slips can be automatically printed),
one or two book carts onto which checked in items are placed in rough sort
order, and a tote for items transiting to another location. Pages may sort the
book cart contents a little finer and then the items are reshelved. If a checked
in item is chosen to fulfill a hold, a hold slip is printed immediately. If the hold
is for pickup at another branch, the item is placed in the tote. If the hold is for
pickup at the local branch, the item is stacked separately. The holds stack is
eventually moved to a holds cart for filing on the public hold shelves.
Some returns are taken to the service desks rather than dropped into the
bookdrop (this is not encouraged). Patrons sometimes insist that the person
staffing the service desk check-in their material right away because of the
delay associated with dropping items into the book drop and actually getting
checked in (and off the patrons account).
Interlibrary delivery is handled slightly differently from patron returns because
the entire tote can be checked in at once. Each tote contains either all returns
or all holds. Therefore, when a return tote is scanned, all the items in the tote
can be unloaded onto a book cart without having to scan each item. The same
system of rough sorting is used to ready the items for reshelving. The items on
the book cart may be sorted finer by pages as part of the reshelving process.
When a hold tote is scanned, is the contents are unloaded onto a book cart for
shelving on the holds shelf.
Workflow at Automated Library
At a library with an AMH system, returns are either fed into the sorter by the
customer using a self check-in station or are fed into the sorter by staff using
an induction station. At both stations, an item must be fed in the proscribed
way so that the built-in scanner scans the items barcode. Customer returns
are checked in immediately by the AMH system and a receipt is issued to the

customer. Because of the immediate check-in, customers rarely ask staff at

service desks to check-in material for them.
The sorter is equipped with several sort locations configured with ergo book
carts, delivery totes, or ergo bins. Items ready for reshelving are sorted to the
book carts. Items that trigger a hold are routed to an interlibrary delivery tote
or ergo bin. Items needing to transit back to the owning branch are routed to
an interlibrary delivery tote. The ergo book carts on the sorter are used by
pages when they do the re-shelving. No fine sorting is needed.
Checking in and unloading an Interlibrary delivery tote received by the local
branch can be handled one of two ways: using the sorter to rough sort or rough
sorting by hand to shelving carts. In the case of KCLS, staff has determined
that it is preferable to use the AMH sorter to rough sort material to shelving
carts which doesnt take advantage of the tote check-in feature. So, even
though each item still has to be removed from the tote, it eliminates the step
of placing the items on shelving carts in order. By inducting them directly into
the sorter, the items are checked in and sorted directly to the appropriate ergo
book cart.

Methodology
In order to identify the effects of using an AMH system, it is important to
compare live operations rather than rely on a sequence of performance tests
that are set up to measure the speed of tasks (e.g. the time is takes to checkin and shelve 100 interlibrary delivery items, the time it takes to check-in,
rough sort and shelve 100 items from the book drop bins, the time it takes to
feed 100 items into the sorter and get them back on the shelves, etc.) The
problem with such tests is that it requires staff to do things in ways they dont
normally do them.
The Problem with Using Performance Tests
If we were to limit the timed tests to some number of items (e.g. 100 items in
the bookdrop), we would be able to get a per item time but it would be
based on a sequence of events that never really happens in the library. Items
dont get processed from start to finish by one person. And they are not
batched in any set number of items (e.g. 100 items). One person may check-in
the items in the bookdrop and then leave them on the carts at the check-in
station. Yet another person might do the actual re-shelving. Not only do
different people do different tasks, they also do them on their own schedule.
One task doesnt necessarily lead directly to the next without wait time in
between.
Testing the speed of each discrete task (e.g. checking in 100 items, rough
sorting to book carts, checking in and labeling holds) creates other problems
for the kind of evaluation attempted here. One problem is that during the
timed period, people tend to perform optimally. The other problem is that wait
time between tasks goes unmeasured. Timing someone checking in 100 items or
even one days worth of daily delivery isnt likely to provide an accurate per
item time because that person will be more focused on their work during the
timed period than they would normally be, and they will want to do their best.
This is generally referred to as the observer effect. In psychology, it is known
also as reactivity.1
Waiting is one of the seven wastes or muda2 that are identified by the Toyota
Production System (TPS) and Lean (http://www.lean.org). Both TPS and Lean
are systems that work to optimize workflows in manufacturing as well as other
types of business operations. When measuring the time a specific task takes,
1

Heppner, P.P., Wampold, B.E., & Kivlighan, D.M. (2008). Research Design in Counseling (3rd
ed. ed.). Thomson. pp. 331.
2
See the muda entry in Wikipedia for a more detailed description of the wasteful activities
that do not add value to a production system as defined by the Toyota Production System.
Available from http://en.wikipedia.org/wiki/Muda_(Japanese_term).

the wait time is eliminated. This is one of the most time consuming aspects of
an inefficient workflow. Because automation removes much of the wait time
associated with materials handling, it was important to find a way to compare
the operations without eliminating wait time from the equation.
Compare Similar Live Operations
Therefore, rather than setting up scenarios that made it possible count steps or
measure time but which altered the natural flow of the work and removed the
waste of wait time, we decided to find a way to compare live operations. We
considered comparing before AMH and after AMH scenarios at one library
but found that the data required to make useful comparisons wasnt available.
Luckily, we were able to identify two very comparable libraries, one which had
implemented AMH and one which was still operating manually.
Kent and Federal Way were chosen because of similarities in circulation,
delivery volume, collection size, and physical size. An AMH system composed
of a sorter with 15 sort locations (8 book carts, 6 totes, one exception bin), two
automated check-in stations and one staff induction station was installed at
Federal Way in December, 2006. Kent Library has no sorter or self check-in
system.
Daily Check-ins
Return to Shelve totes
processed per day
Hold totes processed
per day
Size (sq feet)
Collection Size
AMH CKI

Federal Way
1850
530

Kent
1873
476

541

591

25,000
180,000
AMH since Dec, 2006

22,500
159,000
No AMH

Identifying Relevant Factors

Introduction of AMH dramatically changes how a library works including how
staff are deployed, how quickly material is back on the shelves and available to
customers (turnaround time), and how staff interact with customers. For
example, with an AMH system that includes self check-in, the number of times
customers ask to have their material checked in at the desk is practically
eliminated because all returns (using the self check-in system) are immediately
checked it.
When material is backlogged, staff are asked to perform certain tasks for
customers that dont come into play when all material is checked in and
reshelved the same day. For example, when an item is returned that triggers a
8

hold, the customer gets notified that their hold is ready. If the back office
staff do not get the hold up on the hold shelf before the customer comes to
pick it up, it requires frontline staff to go to the backroom to find the
unshelved item. Sometimes an AMH system eliminates backlog and problems
like this go away; however, an AMH system improperly sized or staffed may not
eliminate 100% of the backlog so these problems may still occur.
In order to identify as many pertinent effects as possible, we compared how
quickly Kent and Federal Way were able to get items back on the shelves and
available to customers (e.g. slow turnaround results in more library resources
being unavailable), and we tried to identify all the ways that the two
operations differed in the time demands on frontline staff, backroom staff, and
customers.
Unavailable Resources
When library resources are sitting in the backroom on a book cart, in a tote or
waiting to be processed, they are unavailable to customers. Unavailable
material represents wasted resources: items that have been purchased by the
library but which the customer cant use. While we can all agree that taking a
long time to get something back on the shelf is not a good thing, it isnt easy to
assign a dollar value to that delay. In order to provide a useful measure of the
effects of the AMH system, it was important to develop a formula that resulted
in a monetary cost per day.
In order to define a daily value of a library resource (e.g. book, CD, DVD), the
acquisition cost was established (the cost purchase3 the item plus the KCLS CMS
Department costs4). The acquisition cost was then divided by the number of
days the item remained in the collection5.
The result established an average daily cost per item of two cents ($.02).
Having determined the daily value (cost) of a resource, it was next necessary
to compare the situations where Kent and Federal Way differed in how long an
item was unavailable. Because the calculation was done on a per day basis,
a day increment was used. This means that if material was returned and
3

Average purchase cost came to $28.24 per item. This includes the cost of Easy Fiction, Board
Books, Easy Nonfiction, Childrens Fiction, Adult, Adult CD, DVD, and CD purchase costs plus
the average vendor processing costs associated with each material type (e.g. DVDs cost $23.26
each to puchase plus an additional $2.19 in processing).
4
KCLS processing costs represents the time spent in technical services to both select and
acquire the book. It also includes the cost of doing additional processing once the book
actually arrives. This cost was set at $3.49 per item and was calculated by determining the
total cost of CMS staff divided by the number of new items per year ($5.81), then dividing
number by the percentage of the departments time that is actually spent on these tasks
described above (60%).
5
KCLS determined that, on average, an item remains in the collection for 5.5 years.

reshelved within one day, it was not counted as unavailable but if it took two
days to get it back on the shelf, the cost applied was $.02 per item per day.
Two situations resulted in material being unavailable: when items are returned
and checked in, but dont make it to the shelves; and when items are returned
but not checked in.
Frontline Staff Time
In order to compare the effect of AMH systems on frontline staff time, three
incidents were identified that affect frontline staff and relate to the AMH
system. The three incidents are Claims Returned, Find Items on Book
Cart, and Request Check-ins.
Claims Returned are incidents in which a customer claims to have returned an
item but the catalog does not indicate that it has been checked in. These
types of incidents are usually handled by Library Assistants who work at an
average hourly rate of $26 (including KCLS overhead.) Each such incident takes
25 minutes6 of the Library Assistants time. Therefore, each Claims Returned
incident costs $10.92.
Find Items on Book Cart are incidents in which an item is listed as checked in
but it has not been shelved. Because the item shows as available in the
catalog, customers often come into the library to get the newly available item.
Because they are not shelved, Librarians ($43 hourly rate including KCLS
overhead) must go the backroom and find the item on a shelving cart. Each
incident requires approximately six minutes of the Librarians time at a cost of
$4.30 per incident.
Request Check-ins are situations where customers ask staff to manually checkin items immediately. These incidents are generally handled by Library
Assistants ($26/hour including KCLS overhead) and take approximately one
minute per incident7 at a cost of $.43/incident.
Backroom Staff Time
The two libraries staff their backrooms differently. Kents backroom staff
manually check-in material from book drops and process delivery. Federal Way
staff feed all material into the sorter via the staff induction station. A rover
6

Karen Daniel, KCLS Circulation Supervisor, provided this figure. She broke down the
transaction as follows: discussion with patron - 10 minutes, shelf check at your branch - 3
minutes, calling owning branch or where the item was returned for a shelf check - 5 minutes,
making decision - 2 minutes, contacting patron about decision - 5 minutes.
7
When a customer requests that one or two items get checked in, it may only take a minute
but more often it is a parent with 10-20 picture books that wants them all checked in so a new
20 can be checked outso this estimate is probably too conservative.

10

position replaces carts and totes on the sorter, makes sure nothing jams, and
generally monitors the sorter. All positions are staffed by Library Assistants
($26/hour).
In order to compare the two operations, we compared the weekly cost of all
staff scheduled to perform backroom materials handling functions (processing
book drop returns, processing totes, processing holds, inducting material into
the sorter, check-in, sorting, and sorter rover).
Customer Time
When an item is checked in but not available on the shelf, a customer who is
monitoring the items status in the catalog can make the mistake of coming to
the library to get the item only to find that it isnt really available. Depending
on the customer, they may leave the library frustrated and discontinue using
the library (per a previous KCLS study, 20% of customers do just that), or they
may request staff assistance. In order to measure this impact on customers, we
assumed that the customer spends 30 minutes round trip, and makes the trip
via car (driving 5 miles). Customer time is valued at the Page Hourly rate
($12.68). We further assumed that such an incident occurs with only 5% of
items that are listed as available in the library catalog but are actually still
parked in the backroom. Based on these assumptions, the cost to the customer
(per incident) comes to $11.38.

11

Results of Comparisons
When we compared all the relevant effects described above, including the
cost of unavailable resources, cost of time spent on incidents involving
frontline staff, costs of materials handling staff, and costs of customer time,
we found that there was difference of $228,605 between the automated and
manual libraries (see Appendix: Charts Showing Comparison of Costs.)
In the case of unavailable resources, the manual library left 584 items
unshelved each day (these items were checked in so would show in the catalog
as available). The automated library left 160 such items unshelved (74%
reduction). This results in an annual cost difference of $3,010. In terms of
items that were returned and not checked in (which means they were also not
shelved), the automated library was able to keep up (no backlog), whereas the
manual library left 1400 items per day. The annual cost difference of these
unshelved and unchecked-in items comes to $9,940 per year. The total annual
cost difference of unavailable resources comes to $12,950 annually.
In terms of frontline staff time spent on three kinds of incidents that are
largely avoided when AMH systems are used, we found that the cost difference
was $55,098 annually. The primary difference was in Request Check-ins which
were reduced to 5 per day at the automated library and as high as 315 at the
manual library (cost difference: $47,322). The manual library reported 12
Claims Returned incidents per week and the automated library reported five
for a cost difference of $3,975. The automated library rarely (.5 times per
week) has to send the Librarian looking for an item in the back room (Find
Items on Book Cart) whereas Librarians at the manual library do this 17.5 times
per week for a cost difference of $3,801.
Materials handling staffing costs8 yielded the greatest cost differences. The
automated library schedules Library Assistants 86 hours per week to do
materials handling tasks (68 hours staff induction to sorter and 18 sorter rover
hours). The manual library schedules 142 hours per week in materials handling
(staffing check-in stations, book drop, tote check-in). Annually, the manual
library spends $75,712 more per year on backroom materials handling staff
than the automated library.
Because of the large number of checked in items that do not get shelved at the
manual library, more customers probably go to the library expecting to find a
book on the shelf that is not. Based on our assumptions (customers come in
expecting to find 5% of the items listed as available in the catalog), we

Staff involved in shelving material were not counted in these scenarios because the time it
takes to fine sort and shelve material is unaffected by the AMH system.

12

estimate that 8 automated library customers make a wasted trip each day
while 29 manual customers do so for an annual cost difference of $84,845.

13

Conclusion
Based on the comparisons of two very similar libraries, one with an AMH system
(two automated self check-in stations, a staff induction station, and a sorter
that sorts directly to shelving carts, delivery totes, and bins), the materials
handling staffing costs are considerably higher at the manual library. In
addition, significant amounts of frontline staff time are spent resolving
undesirable incidents that are largely avoided at the library with the AMH
system. Though more difficult to quantify, AMH systems also result in higher
utilization of library resources by decreasing turnaround time, and also likely
reduce situations in which customers come to the library for an available
item that is not yet shelved.
Considering the fact that the KCLS system also provides tote check-in capability
which significantly reduces the amount of time needed for the manual library
to process interlibrary deliveries, this study (which focused on the difference
between an automated KCLS library and a manual KCLS library) suggests that
installing an AMH system at a library without such a central sort operation
would yield even greater savings than those demonstrated here.

14

Appendix: Charts Showing Comparison of Costs

November, 2008
Unavailable Resources
Unshelved Returns
Items unshelved same day

Fed

Kent
160

Calculations Annual Cost

on Difference Difference
584
424

Cost per item per day

73%

0.02

Cost of unavailable resources per day

Difference in cost per year of unavailable
resources
Returned but not Checked In

Percent
Reduction

Fed

Items returned but not checked in per day

$
3,010
Calculations Annual Cost
Percent
on Difference Difference
Reduction

Kent
0

1400

1400

Cost per item per day

Cost of unavailable resources per day
Difference in cost per year of unavailable
resources
DIFFERNCE IN COST OF UNAVAILABLE
RESOURCES AT NON-AMH LIBRARY
PER YEAR

Prepared by Lori Bowen Ayre, The Galecia Group

8.48

Cost per item per day =

costs associated with
acquiring item / number of
days item remains in
collection (5.5 years)

100%

0.02
$

Library is open 355 days a

year. Days open is
multiplied by daily cost of
unavailable resource.

28.00
$

9,940

12,950

Appendix: Charts Showing Comparison of Costs

November, 2008
Frontline Staff Time Wasted
Claims Returned

Fed

Claims Returned per week

Calculations Annual Cost

on Difference Difference

Kent
5

12

Minutes Spent per incident

25

Hourly Rate of LA

26

Cost per incident

Difference in cost per year

Find Items on Book Cart

Fed

Number of times a Librarian has to go to

backroom to find an item on a shelving cart
for a customer (per week)

17.5

Kent reported 11 but

Newport Way actually
counted and got 13 in one
week. So averaged.

10.92

17

Minutes spent per incident

97%

Hourly rate of Librarian

43

Cost per incident

4.30

Difference in cost per year

Request Checkins

58%

$
3,975
Calculations Annual Cost
Percent
on Difference Difference
Reduction

Kent

0.5

Percent
Reduction

$
Fed

Number of times customer requests

manual check-in (per day)

Calculations Annual Cost

on Difference Difference

Kent

310

315

Minutes spent per incident

Percent
Reduction

98%

Hourly rate of LA

26.00

Cost per incident

0.43

Difference in cost per year

47,322

TOTAL FRONTLINE STAFF COSTS

SPENT PER YEAR AT NON-AMH
LIBRARY

55,098

Prepared by Lori Bowen Ayre, The Galecia Group

Sometimes only 1-2 items

per incident, other time 15
picture books so this is
probably too conservative.

3,801

Appendix: Charts Showing Comparison of Costs

November, 2008
Backroom Staff Time Wasted
Backroom Staff Hours
Hours backroom check-in stations are
staffed per week

Fed

Calculations
on Difference Annual Impact

Kent
142

Hours staff induction is staffed per week

68

Hours sorter rover is staffed per week

18

TOTAL backroom MH hours per week

86

142

Hourly Rate of LAs

Difference in weekly cost (LAs) backroom
staff

11 L.A. hours during the week

and 8 L.A. hours on the weekend

Percent
Reduction

TOTAL BACKROOM STAFF COSTS

SPENT PER YEAR AT NON-AMH
LIBRARY

Prepared by Lori Bowen Ayre, The Galecia Group

2,236

3,692

56
$

26.00

1,456

39%

Conflicting estimates:
Deatrice Barber said: On a good
day, circ staff do rover duties 4
hours a day (28 per week?)
Assistant says it is closer to 2 hours
per week.
Comprised at 2 hours per week (5
days) and 4 hours per week (2
days)

75,712

Appendix: Charts Showing Comparison of Costs

November, 2008
Customer Time
Fed

Calculations
on Difference

Kent

Number of items showing "Available" in

catalog but not on shelf yet

160

584

424

Percentage of "Available" items customers

look up and come in to library to get

5%

5%

5%

29

21

Number of "incidents" per day

Customer time per incident (search and
trip to library)

6.98

Driving cost (assuming 5 miles)

4.40

Total customer cost per "incident"

11.38

Cost per day

239

CUSTOMER COST OF WASTED TRIP

TO LIBRARY

Prepared by Lori Bowen Ayre, The Galecia Group

Annual Impact

Percent
Reduction

Assumine page hourly rate

for patron, 30 minutes per
trip plus 3 minutes looking
up item.

73%

Assuming 5 mile trip and

using cost per mile from
commutesolutions.org.

84,845

Appendix: Charts Showing Comparison of Costs

November, 2008
Breakdown of Costs
Total Annual Costs Elimated with AMH
CKI

228,605
Cost

Percentage of Total
Annual Cost of Unavailable Resources

5.7%

12,950

Annual Cost of Frontline Staff Time

24.1%

55,098

Annual Cost of Backroom Staff Time

33.1%

75,712

Annual Cost of Customer Time

37.1%

84,845

Prepared by Lori Bowen Ayre, The Galecia Group