Robotic Process Automation (RPA)

On Entering an Age of Automation of White-collar Work Through Advances

in AI and Robotics
Recent progress in digitization has had significant impact on white-collar work as well as on
products and their related technology. It has been said that in the next 10 to 20 years, 47% of jobs
will be substituted by automated or robot labor. RPA is a particular type of digitization, and its
implementation is absolutely essential to raising a company’s added value in the future. At KPMG,
we assist corporate problem-solving through our support for RPA implementation, making use of
our prodigious industry knowledge of business innovation and RPA implementation cases in both
Japan and abroad.

Overview of the RPA

What Is RPA and Digital Labor?
RPA comprises distinct types: automation of routine RPA can be described as introducing digital labor (digital
office work and automation of advanced intelligent worker) into the corporate organization in order to carry
processes through means such as artificial intelligence out work that only human labor was formerly considered
(AI). The effectiveness of routine office work automation capable of doing, or to augment human labor in carrying
has been operationally validated in not only North out highly advanced work.
America and Europe but also in Japan, and this type of
automation is now in practical use.

What Is RPA and Digital Labor?

Target Effect
• Invoice processing Automating Increased Eliminating
Accounting &
Finance
• Expenses processing routine office work 1
Quality human errors
• Cash flow statement creation
Improve processing
• Personnel data correction 2 Speed speed by several
Payroll & Benefits
• Year-end tax adjustment hundred times

Application • Filling in application forms Increased

3 30-75% drop
Processing • Data deficiency checking Efficiency

Sales-related • Entering customer information Increased Focus on work with

4 greater added
Clerical Work • Name identification processing Functionality value

RPA Classes and Major Scopes of Application

There are three classes in RPA, and at the present, They are projected to reach Class 3 in five years, able to
companies are starting to create sizable impact automate process analysis, improvement, and decision-
through Class 1 implementation. making.

The Scope of This Document

Automation of Some
Class 1 Automation of Routine Work Class 2 Non-routine Work Class 3 Advanced Autonomization
• Automates even routine tasks that were • Automation of exception case processing and • Advanced AI enables automation of process
performed by human labor due to whatever non-routine tasks analysis, improvement, and decision-making
constraints applied, using business rules • Learning based on data analysis (Deep above and beyond work automation.
Traits

engine, image recognition technology, learning)

workflow, and so on. • Unstructured data processing (ex. natural
• Human labor is needed to deal with situations languages)
such as exception cases, though.

 Data collection (crawling)  Data query (unstructured data)  Decision-making

Applicable

 Data query (structured data)  Answer queries utilizing knowledge base  Complex processing
Tasks

Major

 Data entry (knowledge DB, massive data source)  Advanced analytics, etc.
 Validation  Analytics provided through human
 Logins to multiple systems, etc. intervention, etc.
 Creating rules from aggregated past data

Autonomous Natural language

adaptation processing
Data Analysis
(Conceptual)

& Large-scale
Technology

processing Big data analysis

Business Image Unstructured
Rules Engine Recognition Workflow Information Processing
Machine
AI
learning
“Mega macro” with the power to
“Cognitive AI” “Strong AI/Autonomous AI”
freely perform inter-system processing
© 2018 KPMG Consulting Co., Ltd., a company established under the Japan Company Law and a member firm of the KPMG network of independent member firms affiliated with
KPMG International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.
Why RPA, Why Now?
Surge in Labor Cost in Developing Nations and Evolution of Digital Technology
Since the 1990s, an increasing number of firms have sought to reduce their operating cost by
outsourcing some of their white-collar work to developing countries with lower cost of labor, such
as China and India. However, recent years have seen an upsurge in labor cost in developing
countries, making it more difficult to reduce costs by leveraging labor cost differentials. The
troubling rate of worker turnover in developing nations poses an additional challenge to securing
sufficient level of quality.
On the other hand, technological development has resulted in greater automation in some
industries such as the automotive sector, where the scope of automation has already been
expanded for practical use. Automation of white-collar work through RPA, using technologies such
as AI and robotics, is now poised for acceleration.

Major Impact of RPA implementation

Contribution to Quality Improvement Contribution to Efficiency Improvement
RPA enjoys high affinity with clerical task processing in general, RPA renders human intervention superfluous; work can be
and compared to human labor, is more capable of carrying out carried out in parallel with virtually no resource constraints.
tasks both continuously and reliably. Consequently, errors and The result is that while cost-reduction achieved through
clerical mistakes are greatly reduced, leading to improved task business process outsourcing (BPO) and other measures
quality. based on labor cost differentials and standardization is thought
to top out at 15-30 %, RPA enables cost reduction of 40-
Contribution to Speed 70 %.*
Data editing, screen operation, and other tasks via RPA do not
involve physical manipulation of devices such as keyboard and * According to results for Company A in their final report.
mouse, and therefore a task can be accomplished at tens and
hundreds of times the speed of human labor, resulting in
overwhelming increase in speed of operation. In addition, the
speed of implementation and response to changes, and overall
speed in producing results, will also improve.

Reference: Customer Feedback from Early Adopters of RPA

The efficiency of our processes improved due to RPA implementation, leading to improved quality
while also enabling us to focus more resources on work with greater added value such as improving
the process itself or generating innovation.

Goals & Effects of RPA Implementation Results of Study on Effects of RPA Implementation

1 Improved Quality Response

No. Item Thought to Have the Largest Impact
Rate
• Capable of more continuous, reliable, and high-
quality task implementation than human labor. (1)
1 Decrease in error rate 21% Improved
Quality

2 Speed (1)

• Overwhelmingly greater speed of implementation

2 Improved quality of routine tasks 21% Improved
Quality
and response to changes, speed in producing
results, and speed of operation.
Improved speed of business process (2)
3 19%
implementation Speed
3 Improved Efficiency
• Because a large number of tasks are automated, Reduced dependency on multiple (3)
work can be carried out in parallel with virtually
4 14% Improved
systems and screens Efficiency
no resource constraints.
Increase in STP (straight through (3)
5 11% Improved
processing) Efficiency

Improved Functionality
4 (4)
(Shift Towards Business with Greater Added Value) Accumulation of data for process
6 7% Improved
• Humans are freed from routine work, able to improvement Functionality
engage instead in work with high added value.
• One incidental effect is that various historic data (4)
can be amassed, contributing to identification of 7 Potential reduction of excess tasks 7% Improved
Functionality
further opportunities for improvement.
Source:
“Insights from Early BPO Adopters of Robotic Process Automation,” Horses for
Sources, February 2015.

© 2018 KPMG Consulting Co., Ltd., a company established under the Japan Company Law and a member firm of the KPMG network of independent member firms affiliated with KPMG
International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.
KPMG Standardized Approach
RPA implementation has the potential to change the way processes and business are done. Greater effectiveness can be unleashed
by incorporating business process reform into the perspective rather than simply viewing the matter as RPA tool implementation.
Processes: 2-3; duration: 6–8 W Processes: 10+; duration: 3 M Processes: 30+; duration: 3–6 M Processes: 100+; duration: 6 M+
PDCA/Ongoing
Future Plans & PoC Implementation Launch/Standardization/Upgrades
Improvements
Building Operation Upgrading Operation
1 Future Plans 5 Management Functions 8 Management Functions 9 Monitoring & Improvement
• Design, building, implementation,
• Large-scale RPA management
and testing of RPA implementation
• Formulation & implementation of strategy & horizontal deployment • Regularization of PDCA cycle
& operation processes, rules and
RPA rollout plan policy drafting/implementation • Operational transformation utilizing
organizational structure
• ROI analysis and assessment • Robot analysis and assessment RPA
• Building of RPA integrated
(KPI) system design/promotion
monitoring infrastructure

2 PoC 6 Building and Promoting RPA Operations

• Technology verification and vendor
• Work analysis & selection, requirements specification
selection
• RPA design, building, implementation, version upgrade
• Applicability testing
• Release, operation, maintenance, response to inquiries
• ROI estimate
Preparing to Launch the RPA Organization/Human Resource
3 RPA Organization 7 Promoting RPA Organization/Human Resource Development 10 Development
• Formulation of RPA management • Cultivation of personnel for development and operation work, implementation of • Growth and expansion of personnel
organization & system skill transfer from experts for development and operation work
• Cultivation of personnel to launch • Cultivation of personnel to expand the scope of RPA implementation and • Offshoring and creation of business
RPA implementation promote RPA on an ongoing basis process center

4 Program & Project Management

• RPA program scope, schedule, risk, and resource management
• Integrated change management such as impact survey and regularization of various activities involved in RPA implementation

Building Operation Management Functions

Six different aspects of RPA operation management architecture must be considered during the post-PoC, implementation
launch and standardization phase.
* The following six perspectives are RPA-TOM (Target Operating Model), a KPMG’s consulting methodology.

Personnel & Skills Process

■ Securing personnel who will be ■ Implementation, operation and
responsible for operation Personnel & management processes and
Skills
Process
management functions and functions needed to achieve
developing their skills. maximum effect in a stable
manner.
Outsourcing RPA Organization Organization & Governance
Out
Operation & &
■ The scope of outsourcing to sourcing Governance ■ Organizational makeup and role
Management
third parties in operation assignment for achieving
management. operation management functions.

Performance Technology
Performance Technology
■ Effectiveness evaluation indices ■ IT tools and platform for
for each operation management efficiently achieving operation
function. management functions.

Building and Operating RPA

KPMG’s RPA development and quality standards distill best practices nurtured through a wealth of productive experience to
successfully build efficient and high-quality RPA that leverages the characteristics of RPA tools.

RPA Building Requirements

Process Planning Operation Effect Achieved
3 ４
Specification

1
Risks & Challenges Resolution through Development Standards
2
1 Decline in ROI due to pursuit of
Effectiveness  Leveling of granularity in RPA creation
perfect automation & Quality
Assurance  Increase maintainability
2 Erratic quality and serviceability due
to architecture that increasingly
depends on individual skills Operational  Linkage with subsequent work
Stability
 Detection of any abnormalities
3 Risk of work stoppage due to
Maintenance
that occur
unstable performance
 Avoidance of security risk
4 Compliance risk due to absence of Compliance  Recording and storage of results from
activity trail logging activity (audit trail)

© 2018 KPMG Consulting Co., Ltd., a company established under the Japan Company Law and a member firm of the KPMG network of independent member firms affiliated with KPMG
International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.
Case Study: Accounts Payable Operation Automation
Automaker B has implemented RPA into accounts payable operations at its global shared service center, automating the process.

Humans perform a series of tasks manually.

Implementation

Read invoice Comprehend content Post to general

Verify entered data
Before RPA

information and enter information ledger

• Implemented ERP
system, but
automation stalled.

Order & delivery

system

Robot performs entry, verification, and posting of invoice information.

Implementation

RPA RPA RPA

Read invoice performs verifies performs • Reduction in time
After RPA

information entry data posting spent on routine

office work (approx.
65–75% drop)
• Improvement in task
performance quality
(error reduction)

Observations for the Age of RPA & Digital Labor

Exploring Applicability for Your Firm Knowledge of the Industry As Well As Digital
In a few years, RPA will be implemented around the world. It Transformation
behooves companies to prepare by using PoC on some tasks KPMG supports optimization of deployment of digital labor in
to verify applicability or by conducting researches on business operations, based on knowledge of digital
technology trends, before their productivity is left behind on transformation demonstrated by our trailblazing RPA
the global stage. implementation cases in and outside Japan, and our immense
→ There is ample room for exploration even when enterprise knowledge of business process reform for enterprises in
resource planning (ERP) refurbishment ROI cannot be met, every industry.
since compared to ERP, RPA implementation can be Commitment
extremely low-cost depending on scope. We at KPMG consider it our top priority to build a long-term
Review of Existing Contracts relationship with clients based on trust. We give our all to
BPO and shared services present possibility of even greater provide a solution for the client company that is efficient,
efficiency. If your company has a long-term contract with effective, and one-of-a-kind.
partners such as BPO vendors, we think that a review of the
contract going forward can yield even greater results.

Advantages of KPMG
Our Global Network
At KPMG, we use our global network, a wealth of experience
with cross-border projects and our knowledge of trailblazing
cases on a global level to provide advisory services, such as
RPA diagnosis (status analysis & support in identifying RPA
target scope), support in defining visions for the future,
solution selection support, RPA implementation support, and
process reform support. KPMG Consulting Co., Ltd.

Developing and Utilizing Personnel kc@jp.kpmg.com

All predictable work is possibly replaced with RPA, but on the kpmg.com/jp/kc
other hand, there is concern that Japan will suffer from
serious labor shortage in the future. One of the keys is to give The companies we can provide with the services introduced in this brochure
renewed thought to areas where personnel currently on hand and the scope of such services may be limited from the perspective of the
Certified Public Accountants Act, independence rules and conflicts of interest,
can be utilized, and what to do about education towards that end. among others.

Policy & Governance in Shifting to RPA The information contained herein is of a general nature and is not intended to address the
circumstances of any particular individual or entity. Although we endeavor to provide
RPA is capable of doing anything that can realistically be accurate and timely information, there can be no guarantee that such information is accurate
as of the date it is received or that it will continue to be accurate in the future. No one should
digitized, but applying RPA to everything increases the act on such information without appropriate professional advice after a thorough examination
number of non-optimized processes, retreading the spaghetti of the particular situation.

codes from computing’s past and rendering correction difficult. © 2018 KPMG Consulting Co., Ltd., a company established under the Japan Company Law
For that reason, policy and governance must be set in place and a member firm of the KPMG network of independent member firms affiliated with
KPMG International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.
determining what work is to be automated via RPA and what 18-5035
is to be systematized. The KPMG name and logo are registered trademarks or trademarks of KPMG International.