caterpilar_tunneling.pdf

9B13E030

CATERPILLAR TUNNELING: REVITALIZING USER ADOPTION OF BUSINESS INTELLIGENCE Frances Leung and Murat Kristal wrote this case solely to provide material for class discussion. The authors do not intend to illustrate either effective or ineffective handling of a managerial situation. The authors may have disguised certain names and other identifying information to protect confidentiality. This publication may not be transmitted, photocopied, digitized or otherwise reproduced in any form or by any means without the permission of the copyright holder. Reproduction of this material is not covered under authorization by any reproduction rights organization. To order copies or request permission to reproduce materials, contact Ivey Publishing, Ivey Business School, Western University, London, Ontario, Canada, N6G 0N1; (t) 519.661.3208; (e) [email protected]; www.iveycases.com. Copyright © 2013, Richard Ivey School of Business Foundation Version: 2014-03-31

Caterpillar Tunneling Canada Corporation (CTCC), a subsidiary of the U.S. company Caterpillar Inc., had been experiencing a host of problems — e.g., data inconsistency, uneven reporting and poorly defined processes — with its outdated enterprise resource planning (ERP) system. Between 2011 and 2012, it planned to upgrade to its parent company’s SAP but had to cancel due to resource allocation issues. A business intelligence (BI) platform had been deployed locally as an intermediary solution to facilitate business decision making; after the cancellation, it became an even more critical tool to complement the troubled ERP system in transforming raw and disparate business data into actionable business insights. Faced with limited information technology (IT) resources, an inflexible ERP infrastructure, imbalanced user adoption of BI, and under pressure to generate timely financial and performance reporting to the corporate office, Jon McEwan, the CTCC’s business resource manager and head of the finance department, struggled to turn the existing BI solution into the platform of choice for trusted information distribution throughout the company. The BI platform allowed users to link disparate data sources successfully, but not all business units within the company were ready to adopt the software. As a result, the different levels of participation, technical aptitude and personal motivation created an imbalanced reporting landscape characterized by two types of users. On one hand, the analytics junkies who favoured slicing and dicing interactive datasets on their own and getting hands-on with the latest data visualization utilities. On the other, the canned report users who were passive in performing their own analyses or looking beyond the static results generated, and preferred receiving information from the traditional channels with which they were most familiar. Thus, the BI platform had become both the go-to platform for effective decision making for some and a source of multiple versions of the truth for others. This great divide fostered the emergence of “information insiders” who were more proficient in extracting business insights for decision making than those who were less technology-savvy or less adaptive to new tools and processes. To release the full potential of the existing BI platform, McEwan had to focus his energy on user needs and the flaws of the initial BI platform implementation. A

ut ho

riz ed

fo r

us e

on ly

in th

e co

ur se

C M

P T

6 41

I nf

or m

at io

n T

ec hn

ol og

y fo

r M

an ag

er s

at U

ni ve

rs ity

C an

ad a

W es

t t au

gh t b

y A

dr ia

n M

ite sc

u fr

om J

an 0

5, 2

01 7

to M

ar 2

3, 2

01 7.

U se

o ut

si de

th es

e pa

ra m

et er

s is

a c

op yr

ig ht

v io

la tio

n.

Page 2 9B13E030

SAP BUSINESS TRANSFORMATION AT CATERPILLAR From 2005 to 2006, Caterpillar embarked on a business transformation project to implement SAP across its machine and engine businesses to replace a host of legacy systems and to standardize an enterprise- wide ERP platform.1 To facilitate the transformation, Caterpillar established a special taskforce called MACH 1 that would work with new target deployment sites to localize the Caterpillar SAP template to the site’s business and legal requirements, support transformation to the enterprise business processes, map and load legacy system data into SAP, participate in system user acceptance testing and train local users.2 In 2010, CTCC was selected as one of the sites to undergo SAP implementation. A team of key employees was selected to act as subject matter experts and collaborate with the corporate team and external consultants to assess CTCC’s business processes and determine the gaps that the corporate SAP template would need to overcome. After more than a year of localization sessions and data cleansing efforts by the local, corporate and external teams, the project was put on hold indefinitely, mostly because the implementation was demanding more local resources from CTCC than previously expected. The discontinuation of the project allowed corporate resources to be offloaded to other sites with higher priorities for SAP transformation. Furthermore, the gap analysis revealed that CTCC’s unique concurrent engineering and project-based business model would require a SAP template to be designed from scratch to facilitate the management of the end-to-end cycle from design to delivery for tunnel boring machine (TBM) projects, an endeavour unfamiliar to the corporate and external teams. CTCC was left with no choice but to operate under its original ERP system with no plan to invest further in ERP development. CATERPILLAR TUNNELING OVERVIEW Founded in 1972 as Lovat Inc., CTCC was a leading tunneling technologies provider for soft ground to hard rock geological conditions, specializing in the custom design and manufacture of tunnel boring machines (TBMs) used in the construction of transportation and utility tunnels ranging from two to 12 metres in diameter. The company also offered refurbishments and reselling of TBMs, technical and field services support, product support and spare parts fulfillment. CTCC was headquartered in Toronto, Canada and employed a staff of 330 people, with sales agents stationed in worldwide locations including Russia, the United Kingdom and China. It provided management on transit projects globally and contributed to infrastructure building such as sewers and underground cables in places as diverse as China, Egypt and Colombia.3 Mining companies also use TBMs to create mine access. Some of the high profile TBM contracts that have been awarded to CTCC include the provision of four TBMs for mining an 8.6-kilometre extension of the Toronto Transit Commission’s (TTC) Yonge-University-Spadina subway from the Downsview Station terminal to the Vaughan Metropolitan Centre at Highway 7 and the provision of a TBM to Thiess Tunneling for mining a 3.2-kilometre cable tunnel under the heart of the central business district in the city of Sydney for Ausgrid, the largest electricity distributor in Australia.

1 Vertex, “Caterpillar’s Strategy for World-Class Value Added Tax Management,” May 2012, www.vertexinc.com/ResourceCenter/WhitePapers/PDFs/vertex-vat-caterpillar-white-paper-dsb-review-v2.pdf, accessed May 8, 2013. 2 Caterpillar, “Information Technology,” www.caterpillar.com/cda/layout?m=629355&x=7&id=2492888, accessed May 8, 2013. 3 Kim Laudrum, “Lovat Has Tunnel Vision,” Plant, February 18, 2008: pp. 17—19.

A ut

ho riz

ed fo

r us

e on

ly in

th e

co ur

se C

M P

T 6

41 I

nf or

m at

io n

T ec

hn ol

og y

fo r

M an

ag er

s at

U ni

ve rs

ity C

an ad

a W

es t t

au gh

t b y

A dr

ia n

M ite

sc u

fr om

J an

0 5,

2 01

7 to

M ar

2 3,

2 01

7. U

se o

ut si

de th

es e

pa ra

m et

er s

is a

c op

yr ig

ht v

io la

tio n.

Page 3 9B13E030

In 2008, Caterpillar Inc. announced the entry into the tunnel boring business through the acquisition of Lovat Inc. At the time of acquisition, Caterpillar intended to leverage its global purchasing organization, its global reach and its experience in large-scale manufacturing processes to support and invest in Lovat’s product line and in the tunnel boring business. Lovat was later renamed the Caterpillar Tunneling Canada Corporation and became a part of Caterpillar’s Diversified Products Division, alongside forestry, paving, work tools, original equipment manufacturer (OEM) solutions and defense, and related federal products. As a privately owned company for more than 34 years, Lovat operated as a family business and lacked the organizational governance characterized by a multinational corporation like Caterpillar. Many activities were carried out in a reactive manner. Some decision making was also dominated by undocumented tribal knowledge. As CTCC integrated more closely into its parent company with operations that were highly methodical and procedural, the need to streamline and standardize some of the existing processes grew. At the same time, the parent company realized that CTCC’s concurrent engineering and project-based model was vastly different from Caterpillar’s repetitive manufacturing model. Therefore, not all established manufacturing protocols could be directly applied to CTCC. One of the biggest reforms was in how project management was facilitated. The core of CTCC’s business was to deliver custom tunnel boring solutions to its clients. Every TBM was unique and had varying project requirements including the geological condition of the ground, the size of the tunnel, restrictions to site access, sensitivity of the buyer to the overall cost, etc.4 All these factors could affect the type of machine selected and its configuration. Prior to the acquisition, the project management process was rather lax and lacked the consistency and rigour of regular status reporting. It could be months after the project was completed before the project management team realized that the project was significantly over budget. To ensure that the different projects with varying scopes that were running in parallel were all delivered on time, with the highest quality and on budget, a standard project management methodology needed to be established for tighter control on project scheduling, assignments of tasks, reporting of statuses and identification of bottlenecks. An external consultant was brought in to assess CTCC’s project management needs and facilitate a series of training programs on project duration planning and control, resource and cost management, team communications and dynamics, as well as project risk management. A proprietary web update system from the consultant was implemented to enable fast and reliable updates from employees on their assigned tasks. Marked improvement was observed on delivery times and team communications. BUSINESS SYSTEMS LANDSCAPE A number of companywide information systems had been implemented over the years to meet the expanding needs of business users. There were two predecessors to the existing integrated ERP system. The earliest was a homegrown Microsoft Access database that managed engineering parts creation and basic work order processing. It was followed by a custom business system with additional functionalities designed by a contracted software company. Since not all historical data was migrated to the latest ERP system, the predecessors remained accessible to selected users for the sole purpose of information retrieval. Aside from the ERP system, there was also a custom quality assurance (QA) system for keeping track of nonconformance records (NCRs). The QA system had very basic functionalities for the creation of NCRs and the assignment of advocates for resolving them. It did not provide reporting functionalities.

4 Rick P. Lovat, “TBM Design Considerations: Selection of Earth Pressure Balance or Slurry Pressure Balance Tunnel Boring Machines,” International Symposium on Utilization of Underground Space in Urban Areas, Sharm El Sheikh, 2006.

A ut

ho riz

ed fo

r us

e on

ly in

th e

co ur

se C

M P

T 6

41 I

nf or

m at

io n

T ec

hn ol

og y

fo r

M an

ag er

s at

U ni

ve rs

ity C

an ad

a W

es t t

au gh

t b y

A dr

ia n

M ite

sc u

fr om

J an

0 5,

2 01

7 to

M ar

2 3,

2 01

7. U

se o

ut si

de th

es e

pa ra

m et

er s

is a

c op

yr ig

ht v

io la

tio n.

Page 4 9B13E030

The data captured by the QA database was virtually untapped in spotting trends of the origins of nonconformance incidents. The latest ERP system consisted of modules including project management, sales order processing, item master data, finance, material purchasing, production planning, inventory management, shipping and receiving and time and administration. All transactional business data were captured by this system. The implementation of the ERP system was completed in 2005 and was carried out by an ad hoc team of handpicked employees who were knowledgeable about the company’s operations. The specific ERP package was selected by the executive members of the company, primarily due to the attractive offer from the software company that was at the time expanding its business in North America. While the original ERP package was designed for the European market, the software company wanted to partner with Lovat to launch a pilot program that aimed at customizing the ERP package to suit the needs of Canadian companies. For example, the taxation configuration had to be tailored for the Canadian market. Lovat also requested other parts of the ERP system to be customized to conform to its business processes. The result of the extensive customization was unpredictable system behaviours that the end users and the local ERP support team had to bear, resulting in prolonged dissatisfaction and complaints. ERP implementation had since been a touchy subject in the company. The cancellation of the SAP implementation further intensified employees’ dismay. There were other silo applications or utilities developed and used by departments for their own internal processes that the ERP system was not adequate to support. For example, the TBM sales department established its own quotation tools for TBM contracts, and the power and controls systems engineering unit created its own electronics and hydraulics parts database for facilitating the design process of electrical and hydraulic assemblies. Both utilized data feeds from the ERP system (see Exhibit 1). INFORMATION TECHNOLOGY DEPARTMENT IT was not regarded as a strategic business unit but merely a supportive function at CTCC. Prior to the acquisition, the IT department reported to the operations manager, who was responsible for operations in all production facilities and for maintenance and capital expenditure for buildings and equipment. Typical IT activities included setup and maintenance of computer hardware and peripherals, web and network administration, software installations and upgrades, phone system management, systems security management, IT asset management and helpdesk support, etc. After the acquisition, the IT department temporarily reported to the business resource manager who led the finance department. It was later reorganized and became part of the parent company’s Global Information Services (GIS) unit. While it remained in Toronto, it was under the supervision of Simon Swiss, an IT manager from GIS located in Peoria, Illinois. Its responsibilities also expanded to support a number of Caterpillar facilities outside of Toronto. The reorganization provided an opportunity for CTCC’s IT department to gain a wider exposure to the vast IT resources of the parent company and learn to adopt the standard operating procedures established by GIS. Swiss also became the key contact to help CTCC navigate the corporate IT landscape that was confusing to the local members due to the large number of subdivisions. The most prominent change was on network and web security. By bringing CTCC’s network behind Caterpillar’s corporate firewall, the flow of incoming and outgoing network traffic was secured. Aside from an IT audit performed on CTCC’s computer network, there was no formal assessment performed on the company’s overall information management needs to determine how the subsidiary Au

th or

iz ed

fo r

us e

on ly

in th

e co

ur se

C M

P T

6 41

I nf

or m

at io

n T

ec hn

ol og

y fo

r M

an ag

er s

at U

ni ve

rs ity

C an

ad a

W es

t t au

gh t b

y A

dr ia

n M

ite sc

u fr

om J

an 0

5, 2

01 7

to M

ar 2

3, 2

01 7.

U se

o ut

si de

th es

e pa

ra m

et er

s is

a c

op yr

ig ht

v io

la tio

n.

Page 5 9B13E030

could better leverage corporate’s IT resources. It remained challenging for the rest of CTCC to identify the proper individuals from GIS to engage on specific IT-related matters. Since GIS was a shared service organization that supported all Caterpillar facilities, its response was often not prompt, a situation to which local users needed to adjust to as they were used to the fast turnaround of the local IT team. ERP DEPARTMENT The ERP department was relatively small and consisted of two ERP analysts who managed all facets of the system and the financial application that worked in conjunction with it. This was the only department that had sufficient knowledge on the ERP database structure to perform structured queries for custom information retrieval. Prior to the acquisition, the ERP department was under the same roof as the IT department. After the acquisition, it was put under the supervision of McEwan. Due to the parent company’s unfamiliarity with the local business systems and its inability to advise the ERP team on the future management of these systems, ERP did not become a part of GIS like its IT counterpart despite the fact that many activities that ERP carried out had close ties with the databases and servers that the IT team helped manage. The ERP team was left with managing the ERP and BI systems independently. The separation of the two departments resulted in communication problems. Since they were under the supervision of different authorities, it was sometimes not clear to them in what direction each was heading. Faith Lang was the ERP analyst who not only proposed and oversaw the implementation of the BI platform but also managed the administration and development of the platform. When she first started with the company, it had relatively few reporting capabilities. Canned reports that were supplied as a standard to the ERP system were the only reporting service available to the majority of users. Most of these canned reports were static and not formatted to make the information displayed immediately useful to users. Lang and her colleague were constantly bombarded by requests to create custom queries about the business systems and generate aggregated results. Without a mechanism to store or enable secure sharing of these queries, the same ones were often redundantly created. While in the past the results were delivered to each requestor as embedded queries in Excel files, the future usage and access to these queries that ended up on the users’ machines were out of the staff’s control. The introduction of the BI platform was successful in eliminating most of these problems related to data security, content management, version control and information sharing and somewhat reduced the number of one-off requests. However, the adoption of the technology remained sparse. Lang felt that she was largely accountable for maximizing the benefits that the BI platform could bring to all employees who had the needs to derive insights from raw business data. See Exhibit 2 for CTCC’s organization chart. See Exhibits 3 to 8 for an overview of business intelligence. ORGANIZATIONAL BUSINESS ANALYTICS NEEDS AT CTCC Spare Parts Sales and Field Services As the director of marketing and product support, Norman Neves’s role at CTCC demanded him to grow spare parts sales, oversee field services dispatch and establish the Caterpillar brand in the tunneling segment. When he first started at CTCC, he struggled with getting the data that he needed to make sales projections due to the lack of reporting in the ERP system. Before he came aboard, the finance department reviewed the financials of the spare parts but there was no market segmentation analysis Au

th or

iz ed

fo r

us e

on ly

in th

e co

ur se

C M

P T

6 41

I nf

or m

at io

n T

ec hn

ol og

y fo

r M

an ag

er s

at U

ni ve

rs ity

C an

ad a

W es

t t au

gh t b

y A

dr ia

n M

ite sc

u fr

om J

an 0

5, 2

01 7

to M

ar 2

3, 2

01 7.

U se

o ut

si de

th es

e pa

ra m

et er

s is

a c

op yr

ig ht

v io

la tio

n.

Page 6 9B13E030

performed by the sales team to help drive sales. Although raw data could be manually exported from the ERP system for such an analysis, it was deemed too time consuming. The ERP system did not offer the utilities for the sales team to effectively monitor sales performance by customers, geography and spot trends through time. Neves needed a tool that would allow him to slice and dice the data multi- dimensionally. Neves heard of the BI platform at CTCC, but as a relatively new member to the organization, he was not yet formally introduced to it. He was hopeful that the ERP team would be able to assist him in setting up a reporting solution for market segmentation. With permission from McEwan, Neves and his lead spare parts sales advisor, Stacey Martin, began working with Lang from ERP on requirements. In a few weeks, Lang designed a set of reports and dashboards that allowed Neves’s team to forecast future sales targets and provide feedback to the global sales representatives on what to expect in terms of field services needs. By segmenting historical spare parts sales data by customers, geographical locations, product categories, etc., the department was able to gain insights that no one ever had access to before. They were able to determine the customers that drove sales regions that had the highest demand, during what times of the year sales would likely spike, the types of components with the highest demand, the probability of a quotation turning into sales and more. Neves was able to better communicate with the supply chain team to ensure that inventory was available for the long lead time items and advise the inventory management team on the most optimum stock quantity for specific spare parts. With a good grasp of the types of components that were in demand, he was also able to identify the inactive spare parts in the inventory and devise a disposition strategy. See Exhibit 9 for the interactive report template developed for analyzing spare parts sales performance by geography. Project Management Office (PMO) PMO was the first department to be introduced to the BI platform due to its pressing demand for a better project reporting solution. Before this, project cost reporting was deemed a very challenging task by many. Retrieving the latest project costs from the ERP system could take hours due to the massive bill of materials (BOM) of a TBM. The standard reporting interface from the ERP system could only provide the bulk cost for each project phase, not the details at the task or component level. Visibility to the overall TBM production status was poor considering that each phase could take weeks if not months from design to completion. Although production status for a particular part could be looked up on the ERP system, it did not provide the big picture on where the project was at and what was left to complete. Without a good grasp of what was on the critical path, if ordering a long lead time item was missed and no one caught it, the overall project status could be affected. Lang knew that she had to be proactive to show that she could help. She contacted Nathan Zachery, the PMO supervisor, to get an understanding of what his team needed to monitor projects on a regular basis. Through many iterations of redesign, a project cost template was standardized and used in all control meetings. An interactive report allowed users to retrieve the latest costs accrued for a given project. They could narrow the costs at different levels of a machine down to the nuts and bolts, compare the budgeted versus actual costs, analyze the number of labour hours consumed for designing and manufacturing different parts of a machine, analyze resource loading by reviewing the labour hours each work centre had accumulated, etc. Bound by tight deadlines and high volume workload, not all PMO members were eager to spend time on learning a new set of analytics tools. At the same time, the ERP team needed the project managers’ engagement and feedback to continuously improve and maintain the reporting solution that would work for them. Prior to becoming the PMO supervisor, Zachery had been a project manager himself. He A

ut ho

riz ed

fo r

us e

on ly

in th

e co

ur se

C M

P T

6 41

I nf

or m

at io

n T

ec hn

ol og

y fo

r M

an ag

er s

at U

ni ve

rs ity

C an

ad a

W es

t t au

gh t b

y A

dr ia

n M

ite sc

u fr

om J

an 0

5, 2

01 7

to M

ar 2

3, 2

01 7.

U se

o ut

si de

th es

e pa

ra m

et er

s is

a c

op yr

ig ht

v io

la tio

n.

Page 7 9B13E030

understood the benefits that the BI platform could bring, if all members of PMO could explore the capabilities of the BI platform at the same pace and suggest to ERP the areas that the reporting was not yet able to accommodate. For example, it would be helpful for PMO to see the cash flows of the projects when clients made payments. Since the finance department handled payments in a separate financial system, this information was not visible to PMO unless someone asked for it. As all projects were billed by progress, it would also be helpful for PMO to gain visibility to profit margins for different parts of a machine. Engineering The engineering department at CTCC was made up of a large team of multidisciplinary engineers divided into three units: mechanical design, backup system design and power and controls systems design. To balance resource loading, supervisors needed to monitor the workload of each engineer to maximize resource utilization. All engineers were required to clock their working hours throughout the day based on the tasks they performed. This allowed their supervisors and the engineering manager to determine the number of hours spent on different types of engineering activities such as component design, component testing, research and development (R&D), training and onsite support. The time and administration module of the ERP system was capable of generating reports on the total number of hours committed during a specified period by a particular engineer or by the work breakdown structure of a TBM project. It however could not offer insights on the resource loading by a combination of those dimensions. The BI platform overcame the shortcoming of the ERP system by providing the engineering supervisors with an interactive dashboard that reported on a multidimensional dataset of engineering hours. Users were able to identify the engineers who were working on the same project; the daily, weekly and monthly engineering resource loading patterns by activity types; over- and underutilized resources; the most and least demanding engineering tasks, etc. (see Exhibit 9 and 10). However, user adoption of BI for engineering was slow since it was not a project mandated by management but more of a push by the ERP team to get the supervisors interested in using the tool rather than them actively providing feedback on what would be useful features to embed in the reporting. Quality Assurance Manufacturing a TBM is a complex, concurrent engineering process. Most TBMs were built as one-off units, often hand-built in segments with minimal fixturing by a highly talented workforce.5 Design and planning were critical stages since they had to be done right the first time. Quality was built by experience. It was important for the workforce to learn from past mistakes to continuously improve the design and manufacture process. The premise of the QA system, which was implemented long before the deployment of the latest ERP, was to capture information on all nonconformance incidents for analysis and control. It was a silo system that was managed independently. Its serious flaw was that it did not come with an automated notification feature for driving the workflow. Employees relied on each other’s due diligence to ensure that the NCRs were attended to and resolved in an efficient manner. The QA system also did not have reporting capabilities so it was difficult for the QA team to monitor the overall state of the NCRs.