AI, Labor, and Economy Case Studies

Tata Steel Europe

Contents
Preface 4
Background and objectives 5
Methodology 5
Definition of terms 6
Advanced Analytics at Tata Steel Europe 6

01 Introduction 7
1.1. Background on Tata Steel Europe 9
1.2. Background on the steel industry 10

02 Business context prior to Advanced Analytics (AA) 12

2.1. Business challenges facing TSE 13
2.2. Context prior to Advanced Analytics (AA) implementation 14

03 Move to Advanced Analytics 16

3.1. Decision to implement AA 17
3.2. Management team’s view of Advanced Analytics 18
3.3. Evaluation of AA initiatives 20
3.4. Setting goals and KPIs for AA initiatives 20

04 Development of AA initiatives 21
4.1. Breadth of AA initiatives 22
4.2. Training teams and working model 23
4.3. Data sources 23
4.4. AI techniques used 24
4.5. Challenges in AA implementation 24
4.5.1. Infrastructure challenges 24
4.5.2. Data quality challenges 24
4.5.3. Change management challenges 25

02 page title
Contents
 05 Observations 26
5.1. Business and productivity impact 26
5.1.1. EBITDA impact with high ROI 26
5.1.2. Varying benefits across use cases 28
5.1.3. Raw material savings and yield improvements 28
5.1.4. Rapid model development, longer implementations 28
5.2. Workforce impact 30
5.2.1. History of physical automation at TSE 30
5.2.2. Workforce size 32
5.2.3. Shifts in workforce tasks 32
5.2.4. Workforce training 33
5.2.5. Changes to hiring dynamics 33
5.2.6. Potential medium- to long-term impacts on workforce 33

06 Reflections and implications 35

6.1. Reflections 36
6.1.1. Change management 36
6.1.2. Workforce impact 37
6.1.3. Business and productivity impact 37
6.2. Conclusion 38

07 Appendix 39
7.1. Questions for future research 40
7.2. Exhibit 2. Business summary of the IJmuiden plant 41
7.3. Exhibit 5. Global and U.S. steel production and direct employment 42
7.4. Exhibit 6. Macroeconomic context of the steel industry in Europe 43

03 Contents
Preface

04 Preface
Background and objectives
This case study is a part of a compendium of have not yet produced clearly measurable productivity
ongoing research by the Partnership on AI (PAI) gains for the global economy.1 At the same time,
investigating the impact of artificial intelligence (AI) one major question for the public and policymakers
technologies in the workplace. The objective is to has been AI’s impact on the workforce, both in the
illustrate the tradeoffs and challenges associated changing nature of work and net job loss or creation.
with the introduction of AI technologies into business Our hope is to help ground the conversations around
processes. Through this series of case studies, productivity and workforce impact in examples of real-
we intend to document the different types of AI world AI implementation while highlighting nuances
techniques implemented, as well as discuss the real- across sectors, geographies, and type of AI techniques
world impacts of AI on labor, the economy, and society used. This case study specifically looks at how an
broadly. Researchers often struggle to understand incumbent steel manufacturer facing significant
the economic and social consequences of AI and its industry headwinds launched an internally developed
wide-ranging implications for society. For instance, AI strategy to optimize its production processes with
contemporary economists grapple with the question ML models.
of why ongoing AI and broader digitization efforts

Methodology
Subject organizations were recruited from a pool of project managers, and data scientists. Interviews did
100+ candidates that ‘AI, Labor, and the Economy’ not include plant operators (a more implementation-
Working Group members submitted to the case study focused role), production crews, or customers. As a
project team. The final set of organizations prioritized result, the case study primarily reflects managerial
for study reflects a combination of their willingness perspectives.
to participate in the project and the intention to
profile organizations varying in size, geography Representatives from non-profit and for-profit
and industry sector. The following case study was organizations affiliated with the Partnership on AI’s
developed over the course of four months in the Working Group on “AI, Labor, and the Economy”
summer and fall of 2018. The methodology included supported the case study development by conducting
interviews with different stakeholders at Tata Steel interviews, drafting write-ups, and supplementing the
Europe (TSE) who are directly or indirectly involved case with external research or expert consultations on
in their Advanced Analytics Program - an attempt to the industry or macroeconomic dynamics.
bring AI use cases into the operations of its IJmuiden
steel plant in IJmuiden, Netherlands. The interview
subjects included executive sponsors, plant managers,

1
For more, see “Is the Solow Paradox Back?”, McKinsey Quarterly, June 2018

05 Preface
Definition of terms
While we acknowledge that there is no consensus on of a function previously carried out by a human
the definition of terms such as AI and automation, we operator.2 Following Parasuraman et al.’s definition,
would like to explain how these terms are used in the levels of automation also exist on a spectrum, ranging
compendium: from simple automation requiring manual input to a
high level of automation requiring little to no human
Artificial intelligence/AI is a notoriously nebulous intervention in the context of a defined activity.
term. Following the Stanford 100 Year Study on
Artificial Intelligence, we embrace a broad and
evolving definition of AI. As Nils J. Nilsson has Explainable AI or Explainability is an emerging area
articulated, Artificial intelligence is that activity of interest in communities ranging from DARPA to
devoted to making machines intelligent, and criminal justice advocates. Broadly, the terms refer to
intelligence is that quality that enables an entity a system that has not been “black-boxed,” but rather
to function appropriately and with foresight in its produces outputs that are interpretable, legible,
environment. (Nils J. Nilsson, The Quest for Artificial transparent, or otherwise explainable to some set of
Intelligence: A History of Ideas and Achievements, stakeholders.
(Cambridge, UK: Cambridge University Press, 2010).
In this compendium, a model refers to a simplified
Our definition of automation is based on the classic representation of formalized relations between
human factors engineering definition put forward by economic, engineering, manufacturing, social, or other
Parasuraman, Sheridan, and Wickens in 2000: https:// types of situations and natural phenomena, simulated
ieeexplore.ieee.org/document/844354, in which with the help of a computer system.
automation refers to the full or partial replacement

Advanced Analytics at Tata Steel Europe

Within Tata Steel Europe, the AI-related initiatives and regression and decision trees to more advanced
strategy are referred to as the Advanced Analytics techniques including random forests and some testing
Program (also called Advanced Analytics or AA). of neural networks, which perform more complex
Tata Steel Europe considers AA to be a shift from optimization and prediction tasks.
traditional data analytics. It is defined by a number of
technical attributes: a significant increase in variables
used in the analytics; the use of both structured and
unstructured data (e.g., sensor data vs. images, audio);
more predictive (vs. descriptive) analytics; and the
use of more basic techniques ranging from linear

2
Our definition draws on the classic articulation of automation described by Parasuraman, Sheridan, and Wickens (2000): https://ieeexplore.
ieee.org/document/844354

06 Preface
01 Introduction

07 Introduction
 Introduction

Tata Steel Europe, a leading steel manufacturer with operations throughout the
continent, faced industry challenges, volatile earnings, and increasing customer
demand for higher-quality products. The management team was thus exploring
ways for the company to find sustainable long-term profits, while also looking to
build the foundation for the company’s future through a digitization effort and a shift
to more agile ways of working.

Inspired by the increase of artificial intelligence in manufacturing generally, the

leadership of Tata Steel Europe explored whether they could transform their
largest integrated steel plant, in IJmuiden, Netherlands, into a leading model for
manufacturing analytics and boost the plant’s performance. The leaders recognized,
however, that this change would not come without trade-offs, and that successful
implementation would be challenging.

08 Introduction
1.1. Background on Tata Steel Europe

TSE is a wholly owned subsidiary of Tata Steel Limited, The IJmuiden plant is a fully-integrated steel plant
a company incorporated in India with shares listed on covering the full steel production process. The plant
the Bombay Stock Exchange, and itself a subsidiary of employs about 9,000 full-time equivalents (FTEs4) (See
the Indian conglomerate Tata Group. TSE was formerly Exhibit 1) and produces about seven million tons of
known as the Corus Group, which was a merger steel products annually. An estimated 35 to 40 percent
between British Steel and Koninklijke Hoogovens. It of the plant’s workforce is unionized, with many
was acquired by Tata Steel Limited in 2007. workers a part of the plant’s production crew.

Tata Steel Europe is the second-largest steel producer In FY15/16, the average age of an employee at
in Europe, with a focus on steel products for the Tata Steel Europe was 47 years, and 53 percent of
automotive, packaging, and construction sectors.3 Its the workforce was over the age of 50.5 Most of the
main production assets are located in the Netherlands production crew employees have a high school degree
and the United Kingdom. or equivalent and have been with the company for a
relatively long time period (e.g., 10-25 years). While
This case study focuses on Tata Steel Europe’s largest a production crew job does not require university
plant, in IJmuiden, Netherlands, where the Advanced training, the plant also employs workers such as
Analytics Program was launched. IJmuiden is a small process engineers, which may require a university
port city in the province of North Holland with a degree or above. Some employees also have a
population of roughly 30,000. Ph.D. in a technical field, such as physics. (For more
information, see Exhibit 2 in Appendix 7.2).

Exhibit 1: Breakdown of the workforce at IJmuiden plant

Source: Tata Steel Europe

3
Company filings: Annual report, Tata Steel Europe 2018; Tata Steel Corporate Presentation 2017.
4
FTE refers to full-time equivalent employees at the company. The term is a business acronym and is a conventional unit of measure to
compare workloads across different business contexts. Because labor may be undertaken by part-time employees, it is useful to standardize
work amounts across full-time equivalents, rather than total worker counts.
5
Company filings: Annual report, Tata Steel Europe 2018; Tata Steel Corporate Presentation 2017

09 Introduction
1.2. Background on the steel industry

The global steel industry has undergone tremendous technologies, associated improvements in resource
changes in the past several decades. Today, global allocation, and increased competition.9 One study
steel supply exceeds demand, with a surplus of 540 from the U.S. estimates that the person-hours required
million metric tons in 2017, the second-highest in to produce one ton of steel declined 85 percent from
history. 6 The downward pressure on steel prices of 1980 to 2017 due to automation technology (See
this oversupply is just one of the challenges facing the Exhibit 3). While employment in the steel industry
industry, where even small price reductions can have fell by a factor of five, shipments of steel products in
significant impacts on a company’s profit margins and 2005 was roughly at the level of the early 1960s. Thus,
threaten viability. Raw materials prices have also been output per worker grew by a factor of five in the same
volatile in recent years, exposing steelmakers to risks timeframe.10 (For more on productivity increases,
from input costs that they often cannot control. In steel output, and employment, see Exhibits 3 and 4.
addition, industry-wide factors, such as unionization See additional macro-economic Exhibits 5 and 6 in
and government regulation, may affect employers’ Appendix – 7.3 and 7.4).
workforce flexibility or other managerial decisions
during times of challenging industry dynamics.
Exhibit 3: The labor required to produce the
In the past several decades, steel worker productivity same amount of steel has significantly declined
has increased significantly, and employment in the due to automation
steel industry has fallen. Twenty-five years ago, the TSE
plant in this case study produced three million tons
of steel annually with 23,000 FTEs. Today, it produces
seven million tons of steel with only 9,000 FTEs.7 The
industry trends globally and in the U.S. are similar: The
U.S. steel industry alone shed about 80 percent of its
workforce between 1962 and 2005, or about 400,000
employees. Similarly, global employment in the steel
industry fell 39 percent from 1974 to 1988 and another
9 percent from 1990 to 2012.8 Meanwhile, worker
productivity globally has greatly increased due to new

Source: American Enterprise Institute, March 2018

6
European Commission (2018). “Steel: Global Forum takes important steps to tackle overcapacity.”
7
Reported data from management interviews
8
World Steel Dynamics (2014). Core Report. Based on a sample of employment data from global steel producers (sample set accounts for
about 35 percent of global steel production over the period).
9
Collard-Wexler, A.; De Loecker, J. (2015). “Reallocation and Technology: Evidence from the US Steel Industry.” American Economic
Association.
10
Ibid.

10 Introduction
Labor protection laws and workforce participation in labor unions can influence employers’ decisions about
workforce reductions in light of these industry dynamics. The Netherlands in particular has a relatively low level of
labor market flexibility,11 according to a study that looked at measures of redundancy rules and costs for EU and
OECD countries.12 On the other hand, the percent of total employees organized in unions in the Netherlands has
declined steadily in recent years, down from 26 percent in 1995 to 18 percent in 2011, and the Netherlands has
a lower rate of unionization than other EU countries.13 At TSE’s IJmuiden plant, an estimated 35 to 40 percent of
workers — largely members of the production crew — are unionized.

Exhibit 4: US primary metals industry — output and employment, 1990 – 2018

Source: US Bureau of Labor Statistics. Primary metal is category NAICS 331. Iron and Steel mills ferroalloy manufacturing is NAICS3311

11
Labor market flexibility refers to a company or economy’s ability under a certain jurisdiction’s regulations or laws to make hiring, firing, or
workforce condition decisions around its employees. It can be measured in various ways, such as an index.
12
Lithuanian Free Market Institute (2017). “Employment Flexibility Index.”
13
L. Fulton (2015). “Worker representation in Europe.” Labour Research Department and ETUI.

11 Introduction
02 Business context prior to
Advanced Analytics (AA)

12 Business context prior to Advanced Analytics (AA)

2.1. Business challenges facing TSE

Tata Steel Europe’s management faced three key challenges, which contributed to its decision to explore analytics
and AI. These challenges were primarily driven by global and European steel industry trends, as well as by TSE-
specific customer and labor dynamics:

Constant pressure to improve margins in a capital-

1 intensive and competitive industry

Due to the oversupply in the steel industry, steel manufacturers are constantly forced
to improve processes and identify cost savings to increase profit margins that can
be quite volatile. The TSE executive who oversees the Advanced Analytics program
noted, “We are in stiff competition with many players. If you don’t improve or
innovate, you will be at the bottom.” Because of the level of competition, even small
improvements can have an outsized impact. A few percentage point improvement
in yield can have a substantial impact on EBITDA (a common proxy for cash flow)
— and improving EBITDA14 is critical for funding capital-intensive projects that
allow the company to continue operating, such as investing in new equipment or
maintaining old assets.

Raw materials (mainly iron ore and coke, a high-carbon fuel) are the major
components of TSE’s cost base: A TSE technical director estimates that these
materials consume about 63 percent of the IJmuiden plant’s revenues. Labor costs
are the second-largest cost driver for TSE. As one TSE executive put it, “It’s a very
harsh business in steel: very high revenues [but] with very thin margins. We are
affected by a variety of forces — inflation, wages, raw materials costs. It means that
we have to continuously improve.”

Customer demand for higher-quality products

2 Steel is commonly perceived as a commodity product (one in which many

companies’ products would be considered identical substitutes for one another),
yet there is significant customer demand for high-quality steel products that meet
regulatory or other industry-specific product requirements. The IJmuiden plant
alone produces about 200 different steel products, with up to 50,000 different
specifications (stemming from different quality and regulatory requirements) for
customers in sectors such as automotive, construction, and packaging. For example,
an automotive customer can use 20 to 30 different steel products in a car. Higher-
quality products often require higher-quality raw materials. Because of these
dynamics, steelmakers across the globe compete for the highest-quality iron ore
and coke, which increases raw materials costs.

14
Earnings before interest, tax, depreciation and amortization (EBITDA) is a common proxy for a company’s cash flow. This metric can be
sensitive to both company-specific volume changes (producing more steel), as well as industry pricing levels (higher steel price-points), and
thus comparing directly across time periods can have limitations.

13 Business context prior to Advanced Analytics (AA)

 Aging workforce and potentially lower supply of future steel

3 plant labor

The production workforce at Tata Steel is aging: A large portion of plant operators
will reach retirement age in the next 5 to 15 years, which will lead to an increased
workforce attrition rate.15 When operators retire, they take with them years of
valuable experience, which cannot immediately be replaced by hiring new plant
operators. There are a number of reasons that it may not be easy to replace the
soon-to-retire plant operators:

• Cost considerations: In the Netherlands, the unemployment rate has been on

a steady decline since 2014 and is currently below 4 percent. Conversely, wage
growth has been on a steady increase.16 Such cost factors, combined with the
fact that it will take years for a new operator to gain the experience and thus
skills necessary to fully replace a retiring operator, is incentivizing TSE to look at
technological solutions, such as its AA program.
• Job preferences: Finding labor for steel production is challenging in the
Netherlands. Over time, steel plant jobs have also become seen as less desirable,
due to the working conditions and inconvenient working hours (e.g., night shifts).17

2.2. Context prior to Advanced Analytics (AA)

implementation

Tata Steel Europe has faced volatility in its profitability of physical process could prove to be expensive
over the past decade due to increasing competition compared with the variable savings it would provide.
in the global and European steel industry. To address
this, the management team launched a program called TSE’s IJmuiden plant, like many other plants globally
the “Sustainable Profit Program” in 2014 to identify and especially in Western Europe, was built with older
long-term sources of stable profit. equipment and assets, with some machinery dating
back to the 1960s. Installing new equipment is highly
Although automation has been a major driver of capital-intensive, and, given industry trends, the
continuous improvement in steel manufacturing company did not want to rely merely on funding these
processes over the past decades, the potential gains types of projects for long-term sources of stable profit.
from additional automation are perceived to be
increasingly challenging to capture. Today, a majority In 2016, management at the IJmuiden plant began to
of the manual processes have been automated, often explore an Advanced Analytics Program as a potential
with production crews operating machinery from candidate for the Sustainable Profit Program because
control rooms. Automating the remaining pieces of belief it may drive long-term profitability benefits.

15
In FY15/16, the average age of an employee at Tata Steel Europe was 47 years, and 53 percent of the workforce was over the age of 50. Tata
Steel Europe Sustainability report 2017.
16
EUROSTAT, Netherlands, as of December 2018.
17
European Foundation for the Improvement of Living and Working Conditions (2015). “Manufacturing: Working conditions and job quality.”
Dublin: Eurofound Working Conditions Survey.

14 Business context prior to Advanced Analytics (AA)

The consideration of AA was not without drawbacks; management expected that an Advanced Analytics Program
would come with challenges from legacy systems: In European steel plants, roughly 95 percent of the assets are
older than 25 years.18 The technical infrastructure is often outdated as well, leading to data that can be scattered
across plants or various IT systems, and sensors that capture inaccurate or inconsistent values. In aggregate,
reliable analytical insights can be quite challenging and cumbersome to obtain.

Despite the challenges, an Advanced Analytics Program still appealed to TSE management. It could offer
sources of stable and sustainable profit with relatively low investment, while also positioning the company for the
future. Typically, people tend to associate automation with job elimination, but AA offered potential for revenue
generation and cost savings that were not tied to decreasing human headcount (for instance, more efficient use of
energy); this, in turn, could help management maintain strong relations with its labor unions and workforce.

“In the end, it’s all about our customer demand and being able to serve them well,” said the TSE executive in
charge of the transformation program. “We found that we needed to step up EBITDA and address our business
challenges to fulfill this. How can we think of projects that can deliver EBITDA for the short term but also are
sustainable? Out of this search for value, we came across AA as an opportunity.”

18
McKinsey & Co. (2013). “Competitiveness and challenges in the steel industry.” OECD Steel committee.

15 Business context prior to Advanced Analytics (AA)

03 Move to Advanced Analytics

16 Move to Advanced Analytics

3.1. Decision to implement AA

As part of the Sustainable Profit Program, management explored several options to increase EBITDA
(See Exhibit 7):

Option 1: Invest in new equipment, machinery, and plant upgrades. This would require high capital expenditures,
offer a high return, and require a medium payback period (the amount of time needed to recoup the investment).
Option 2: Invest in additional physical automation technology. This would require a mid-sized capital investment
and generate a variable return over a long payback period.
Option 3: Invest in the Advanced Analytics Program. This would require a smaller capital investment and generate
a mid-sized expected return with a short payback period.

Exhibit 7: Comparison of investment options at Tata Steel

1 Payback period refers to the expected time for an investment to recover its initial outlay through generated returns or savings, or the
time to reach a break-even point. Payback period is a simple way to evaluate the risk of undertaking a project.
2 Investment in a blast furnace is a hypothetical example and not one Tata Steel
Source:seriously pursued
Employee at the time
interviews, of interviews,
expert AA. However, theanalysis.
team
investment case is illustrative of the investment size and attributes.

17 Move to Advanced Analytics

These options were not considered mutually 2016, amid this climate, the company evaluated using
exclusive by the management at the IJmuiden plant, Advanced Analytics to improve its manufacturing
and each had their pros and cons. However, given operations. TSE management decided that they
the limited availability of capital, management needed to complete the digitization effort first,
was reluctant to take on highly capital-intensive though, so they would have the technical infrastructure
investments unless they were critical to running plant in place prior to initiating the Advanced Analytics
operations. Additionally, implementing further physical Program.
automation technologies offered limited upside
due to the relatively low expected benefits and long “We tried to demystify digital issues — in the end, it’s
payback period. all about new and old sensors of data, integrating it
into the cloud in a centralized place, and improving
These considerations made the AA program more the models on top of that,” said the TSE executive
attractive than its alternatives. At the time, TSE was who directed the transformation program. “We
engaged in a broader organizational and digitization developed a digital roadmap so that [our employees]
effort to improve operating routines, improve data understand where the pieces fit together in our
capture and quality, and to shift to cloud storage. In broader strategy.”

3.2. Management team’s view of Advanced Analytics

Tata Steel Europe’s management team saw significant • AA could improve product quality: High-quality
opportunities presented by the Advanced Analytics products differentiate the company in a competitive
Program, as well as its limits and potential challenges. market. The program presents opportunities to
improve customer relationships through higher
According to management, launching AA offered a quality products and to increase margins through
variety of benefits: better pricing.
• AA use cases did not focus on labor reduction:
• AA is a low CapEx, 19 high-return investment:
The AA effort was very value-driven and needed
Process improvements and cost savings driven by
to demonstrate direct impact on EBITDA to be
Advanced Analytics can achieve the comparable
approved. In sizing the value of potential use cases
results of alternatives that require large capital
for the AA program, value was most often identified
expenditures, such as investing in a new blast
through yield or throughput improvements rather
furnace (an investment on the order of hundreds of
than labor reduction. Management preferred to
millions of Euros). The TSE executive who oversees
create value through revenue generation or non-
the Advanced Analytics program noted, “It is much
labor operational savings (consistent with no direct
cheaper to invest in AI than to invest in a new
labor reduction achieved in AA’s implementation).
blast furnace. This gives us a chance to achieve
similar magnitudes of impact without large CapEx • TSE has the talent to drive AA implementations:
investments.” Tata Steel Europe takes pride in its workforce and
has rolled out a variety of training initiatives to give
• AA could target the largest cost drivers: Raw
its workforce the skills and tools to develop and
materials are the largest cost driver for the steel
implement Advanced Analytics use cases. TSE’s
plant, and AA presents an opportunity to address
engineer-heavy workforce was highly comfortable
this cost driver given the availability of data related
with analytics and large data sets, based on internal
to raw material processes.
19
CapEx refers to capital expenditures. Capital expenditures are funds used by a company to invest in new physical assets (e.g., plant
equipment) or to maintain existing equipment. These funds can also represent funds to invest in new projects.

18 Move to Advanced Analytics

 assessment, and could thus be trained with the right Once Tata Steel Europe’s management team
toolkit for more advanced analytical techniques. committed to the Advanced Analytics Program,
As one TSE executive noted, “We have a large they were determined to mitigate the cultural and
population who are very connected to this type of technical challenges that might be brought forth
work. We don’t hire a ton of people externally, but by a transformation effort. Planning for change
we train them up from our incumbent population management was a conscious part of the program’s
of process engineers [...] The results have not let us upfront strategy. This planning included two key areas:
down at all.”
• Investment in training programs: Management
was willing to invest resources for (ongoing) training
Still, management clearly sees AA not as a panacea
programs to give a portion of their workforce the
but as one new tool for the continuous process
necessary skills to develop machine learning and
improvement that is critical in the competitive steel
statistical models, to work in agile teams, and to
industry. The management team has looked for use
translate findings into actionable recommendations.
cases with certain features that make them good
The training even included executives and the CEO,
candidates for AA - for instance, data availability and
to demonstrate that a successful implementation of
quality, as consistency across various plant sensors
AA could take place only when the methodology
can be a challenge. Management also looks for a
is understood at all levels. As of this writing,
compelling business case, such as addressing one of
roughly 200 FTEs have been trained through these
the major business challenges or cost drivers (e.g.,
programs (see Section 4.2 - “Training teams and
raw materials savings, process stability improvements).
working model” for further detail).
“When we were in discussions about this strategy
two or three years ago, we saw AA not as a magic • Executive sponsorship: Management knew that
wand, but a practical way to improve our business sponsorship, both in terms of funding and creating
and processes,” said the technical director of TSE’s project owners, would create more organizational
AA program. “We have positioned this program momentum. Executives would not treat the
as something that serves all other improvement program as a side project but as a central part of
programs, such as quality improvement. AA is just one the organization’s strategy.
of the methods that we use.”

19 Move to Advanced Analytics

3.3. Evaluation of AA initiatives

Tata Steel Europe found no shortage of ideas sponsorship from at least one company executive.
when it came to AA initiatives and use cases, which To gauge success of an AA initiative, each use case
constituted a multi-year effort with specific annual had specific metrics for monitoring and tracking
savings and EBITDA uplift targets. progress. Use cases were expected to demonstrate
improvements over a defined period of time, typically
Rather, the challenge was choosing where to focus – from six months to a year. In the initial wave of
and then implementing the initiatives (e.g., developing use cases, prioritization was largely driven by data
models, incorporating change management, and availability and quality, as well as the impact on the
updating infrastructure as needed). The team biggest cost drivers, such as raw material optimization.
evaluated the potential of AA initiatives along a As of late 2018, Tata Steel Europe has implemented
variety of dimensions, including value potential, data about 15 use cases of advanced analytics across its
availability and quality, whether they had sufficient steel manufacturing operations, with 30 more under
complexity to warrant the use of advanced analytical development. More than 200 use cases have been
techniques, implementation feasibility, and buy-in or identified for future phases of the program.

3.4. Setting goals and KPIs for AA initiatives

At Tata Steel Europe, AA use cases are treated as any Each use case is assigned a target to hit. While metrics
improvement project, requiring rigorous performance are often operational KPIs, each use case needs
tracking measures such as key performance indicators to have clear financial benefits and is expected to
(KPIs). A specific KPI is tracked for each use case break even financially in a defined period of time.
depending on its core objective, such as yield The use case project team develops a business case
improvement or reducing the rejection rates for with clear targets, to be presented to an evaluation
product quality standards. For example, one use case’s committee with an executive’s sponsorship. Since
KPI focused on reducing the rejection rate for a quality launching the AA program, some use cases have been
assurance test of a key product. more successful than others, with a few significantly
outperforming the initial target and others realizing
benefits below the initial targets. Overall, however, the
program has exceeded its target across the portfolio
of use cases.

20 Move to Advanced Analytics

04 Development of AA initiatives

21 Development of AA initiatives
4.1. Breadth of AA initiatives

In the initial phase of Advanced Analytics at Tata Steel Europe, the company prioritized 15 use cases, each
focusing on a different process or approach, ranging from energy reduction to product quality improvement to
raw material savings.

Examples include (See Exhibit 8 for more details):

• Recipe optimization for coke20 quality: Model to improve the optimal quality mix of raw materials to reduce
raw material costs with a similar- or higher-quality output of steel products.
• Early warning model to reduce downtime in the steel plant: Predictive model to alert operators when a
major process could become unstable and cause production downtime and productivity losses.
• Product quality improvement: Project to lower the rejection rate on a quality assurance test for finished
products.
Today, most of the AA use cases focus on direct manufacturing improvements. Other corporate functions, such as
supply chain, procurement, sales, finance, or human resources, may be addressed at a later stage.

Exhibit 8: Overview of select AA use-cases across the portfolio

1 The three examples shown are categories of use-cases focused on a particular production process (e.g., quality improvement, recipe
optimization). Some categories include multiple use-cases (recipe optimization = 2 use-cases; early warning signal = 1 use-case; product
quality = 3 use-cases; remaining cases = 9 use-cases).
Source: Employee interviews, Tata Steel Europe data, team analysis.
20
Coke is a high-carbon fuel and a critical raw material input in the steelmaking process.

22 Development of AA initiatives
4.2. Training teams and working model

Around 200 FTEs, including data engineers, data interact with the analytical models. The
scientists, project managers, and company executives, teams worked in an agile21 manner over 10 weeks to
were trained in various analytics courses. Of those, develop iterations of analytics use cases, extracting
about 50 actively developed the use cases and built and assessing data and designing actionable projects.
models. Training was relatively rigorous, in some cases
including up to 70 hours of cumulative training for To strengthen these teams, Tata Steel Europe sought
more technical roles. to find hybrid candidates from within the company
who were knowledgeable about steel operations and
Advanced Analytics teams were cross-functional, about data science to develop machine learning or
including data scientists, data engineers, and subject- statistical models. As an executive noted, “We are
matter experts (such as plant operators), to encourage in the business of making steel. We are not in the
collaboration and ensure that people with different business of developing the best AI algorithms. The
areas of expertise were heard. Teams sometimes objective of the analytical models is to help us make
included user interface designers to build out digital better steel. You need to know both.”
tools, so executives and production crew teams could

4.3. Data sources

As part of transformation efforts, Tata Steel Europe has been shifting its data storage to the cloud, while also
undertaking an effort to increase data centralization, accessibility, and quality. As part of the AA program, the
company has also expanded its data-capture techniques across the IJmuiden plant, including standardizing data
capture and leveraging data-capture techniques for unstructured data such as audio and images.

Most of the available high-quality data was already well-structured as it was sourced from plant sensors.
Therefore, more advanced machine learning techniques, which are uniquely suited to analyze unstructured
data (such as audio or images), might have provided less incremental benefit over less sophisticated analytical
techniques at that point. In some cases, however, Tata Steel has been working on using unstructured data from
“smart” cameras and listening devices. In the early warning slopping22 use case, for example, the team has started
to incorporate images and plant audio to predict if the furnace may overboil leading to production downtime.

21
Agile is a development method, typically for software and design, where cross-functional teams (e.g., data scientists, user experience
developers, business translators, subject-matter experts) work closely together operating in iterative cycles to build a product, in this case
an advanced analytics model and implementation.
22
Slopping, also referred to as “overboiling,” is a technical term in steelmaking when slag foam from liquid steel overflows from the
production vessel and can cause significant delays in the production process.

23 Development of AA initiatives
4.4. AI techniques used

The teams working on the portfolio of AA use cases considered and used a range of techniques — from simple
linear regressions, to Monte Carlo methods, to more advanced random forest models — with the next stage of more
mature projects testing neural networks. The type of technique used depended on the amount of data available,
quality of the data, and the type of problem (e.g., multi-dimensional vs. fewer variables).

The teams discovered that more advanced techniques (e.g., neural networks) did not always generate the best
results. As data availability and quality increase, this could change, making more advanced techniques more viable.
However, the more advanced techniques were often more difficult to develop and implement — and they have not
yet been proven to show significant business impact.

4.5. Challenges in AA implementation

4.5.1. Infrastructure challenges

From a technical perspective, challenges often revolved around data engineering rather than the development
of analytical models. To be successful with more advanced analytical techniques, it was necessary that many
data streams were readily available and were properly cleaned and interconnected. Data centralization and
accessibility was greatly enhanced by a cloud-based IT and data infrastructure that was developed during the
launch of AA initiatives.

At one point, different processes and plants used different database types and servers. The company’s IT
infrastructure could not support advanced analytical techniques due to fragmented batch data storage. Then the
company began its shift to the cloud, centralizing data sources for easy access.” We create so much data in our
production process, but it’s all in scattered databases with a terabyte here and a terabyte there, and we only use
a small part of that,” said the director of the transformation program at TSE. “When you do AA, you need to tap
into the cloud to centralize it all.”

4.5.2. Data quality challenges

The quality and consistency of the data was another major hurdle. Sensors across plant operations could be
decades old or improperly calibrated, and poor-quality or incorrect data could lead to improper or misleading
training of analytics models. The data engineering team needed to invest significant time into cleaning and
structuring the data to ensure that data reliability for the AA use cases. Often, the manual effort of data cleaning
was done on a case-by-case basis.

The major challenges in data quality highlighted the need for TSE to invest in new data sensors across the plant
and to standardize data cleaning practices to lay the foundation for future AA use cases.

24 Development of AA initiatives
4.5.3. Change management challenges

Change management, for both the management team and direct manufacturing staff, was often highlighted as
one of the most significant challenges in implementing the Advanced Analytics Program. When done effectively,
however, it also proved to be one of the most important drivers of implementation success. “People are an
important factor. After we built models, we had to convince the operators to use the models,” said an Advanced
Analytics project manager.

Data scientists implementing models in plants found that plant operators could be initially reluctant to use or
trust the analytics model, and often trusted the “old way of doing things” more than a model that could not be
fully explained (because of the computation-intensive statistical techniques on which it relied). There was also a
sense of pride in the domain expertise that had been previously been developed through years of on-the-job
experience. Yet, according to AA project managers and plant managers, operators tended to be impressed by
the results and would start trusting a model if, for example, it was effective at signaling a potential break-down in
a process. From management’s perspective, the operators did not fear technology in general, but they needed to
build trust with the new technologies and recommendations.23 Explaining the model’s recommendations in a way
that translates to how an operator understands the process proved critical for earning their support.

Data scientists and plant managers cited two strategies that proved effective to get buy-in and to increase the
adoption of the analytical tools:

• Get input early on from domain experts and plant operators: Bringing domain experts, such as plant
operators, into the problem-solving and development of the model early on led to far better results
than having data scientists work separately, both in the model performances and in the buy-in from the
operators. Management reported that, following this process, the operators were less likely to perceive the
recommendations from the AA model as telling them they were doing their job wrong. One plant manager
required data scientists to be in the same room as operators during the model development: “We wanted
[data scientists] not to sit in a room far away, and instead [to] spend two days each week inside the steel
plant,” he said. “It’s necessary to make sure that the operators and our production crew are involved in the AA
program together.”
• Use data visualization tools to help operators understand the model: Analytics project managers also
realized the effectiveness of building user interface and data visualization tools deployed on tablets to help
users more easily understand the rationale of the AI system and interpret the outputs of models. Digital tools
such as tablets could help operators understand the recommendations in language that could be easily
understood and also allowed operators to interact with the results of the models through digital interfaces or
visualizations. “We are thinking more and more of moving from models to tools that other people can use, so it
is more empowering to the people that are in the plant working with it,” said an AA project manager.

23
Employee interviews did not include plant operators, and, thus may not fully incorporate their perspective.

25 Development of AA initiatives
05 Observations

26 Observations
5.1. Business and productivity impact

In its first year, management reported that the Advanced Analytics Program at Tata Steel Europe had a direct
and meaningful impact on profitability (e.g., EBITDA). For instance, management found that the AA program
increased productivity at TSE overall, a key metric in the steel industry, by delivering roughly 13 percent higher
EBITDA with the same production staffing levels. The initial phase of use cases focused on the “low-hanging
fruit” implementations, often driven by data availability and ease of implementation. More advanced use cases
may require increased investment; however, executives expect the future returns to be potentially higher than the
initial phase of use cases as more advanced techniques are used or as AA is implemented in higher
value processes.

5.1.1. EBITDA impact with high ROI

All the Advanced Analytics use cases implemented “For every dollar invested, we received $5 in return,”
in the program’s first year had a substantial EBITDA one TSE executive reported (See Exhibit 9 for details).
impact for the IJmuiden plant, though some use cases
did deliver value beneath their initial target. One The financial benefits took a variety of forms, primarily
TSE executive noted that the AA program overall through direct savings on raw materials and yield
exceeded its initial target. According to a technical improvement, as well as indirect margin improvement
director at Tata Steel, the use cases the AA program from higher-quality products (and the consequent
focused on were “issues we could not have solved shift toward higher prices earned). The investments in
with conventional improvement methods.” He expects the program included setting up and implementing
the benefits to continue. the program (e.g., costs associated with digitization
effort), training employees, and labor to develop and
From managerial perspective, Advanced Analytics implement models. Additional costs include some
proved to be a very efficient use of TSE’s capital and of the IT infrastructure, such as cloud storage; these
resources compared with alternative investments, costs are not insignificant, but also were not major cost
such as new plant equipment (e.g., a blast furnace) or drivers.
physical automation technologies: It required lower
upfront investment and had much higher return on
investment (ROI).

27 Observations
Exhibit 9: Advanced Analytics problem yields an estimated 5:1 ROI

Source: Employee interviews, Tata Steel Europe data, team analysis.

5.1.2. Varying benefits across use cases

The portfolio of AA use cases had a significant impact of these use cases, the quality control rejection rate
on the business, though the size and type of benefits for a product was reduced by 95 percent in just three
varied across the 15 AA use cases implemented to date. months.

For example, one of the most successful categories Due to the company’s thin operating margins (typical
of use case implementations focused on process across the industry), even small operational and
stability improvement and led to roughly 20 percent of productivity improvements (e.g. yield) can have large
total benefits realized from the AA program.24 In one impacts on TSE’s profitability.

Three use cases have been implemented to address process stability. Together, they account for about 20 percent of the total portfolio
24

EBITDA benefits.

28 Observations
5.1.3. Raw material savings and yield improvements

As noted above, the financial benefits from the 15 margin improvements. A focus on product quality
AA use cases implemented to-date come largely improvements has allowed the company to expand
from raw material savings and yield improvement, its book of business on higher-quality and higher-
not through labor reduction. Raw materials savings margin products. This is particularly helpful within
accounted for 50 percent of total program benefits the automotive sector, for example, where quality is
and yield improvement about 30 percent. There have important for manufacturers and is a strict regulatory
been no direct labor cost savings to date. The initial requirement. Higher product quality also leads to
analytics use cases were chosen because they focused higher customer satisfaction, which is critical given the
on IJmuiden’s largest cost drivers, such as raw material excess capacity in the industry and the struggle to sell
costs, rather than labor. a company’s full supply of steel.

In addition to these direct savings, the use cases have

also produced indirect financial benefits through other

5.1.4. Rapid model development, longer implementations

Since beginning the AA program two years ago, the that we can’t keep up with all the strong ideas we have
team at TSE reports they reduced the time required generated by AA,” an AA project manager said.
for analysis and AA model development from about
three months to 30 minutes - yet this is only one The team hopes to expand the impact of the AA
part of the overall process (and model development program into areas such as research and development,
could take longer in novel use cases). In general, TSE specifically about the chemical properties of steel.
found that developing an analytics model takes a “There is going to be much more data coming
fraction of the time associated with cleaning the data out of these initiatives,” said a senior executive at
and implementing a model on existing production Tata Steel Europe. “We are only just scratching the
processes. The AA initiatives have improved the surface. I think we could be doing a lot more to make
efficiency of labor focused on analytical model completely un-thought of types of products.”
development, though other significant enablers
remain in the process. “The challenge for us today is

29 Observations
5.2. Workforce impact

This case demonstrates how one company avoided further reducing labor costs. In the case of Tata Steel
immediate AI-related workforce reductions by Europe, the AA program had less impact on the size
retraining employees and committing to retaining of the workforce than on training, workforce tasks,
them through retirement age. It is not always the case and hiring dynamics — partially driven by the design
that AI implementation results in immediate workforce and intentions of the AA program. Yet even at TSE,
reductions. Instead, introducing AI technologies in the automation-heavy steel industry, core business
may affect the workforce in other ways.25 In some functions remain that may also see transformations
cases, AI can lead to quality improvements that were with AI’s incorporation, such as procurement and
previously unattainable. In others, the goal of using AI finance. As these functions have not yet been subjects
may be to reduce non-labor costs. The steel industry, of AA use cases, there is not case-specific evidence of
even before AI and AA technology, had already workforce impact in these domains.
witnessed a significant reduction in employment due
to automation, which limits the potential savings of

5.2.1. History of physical automation at TSE

Tata Steel Europe has had an industrial automation - for instance, in manually configuring settings on a
program for more than a few decades, while the AA computer or monitoring system). Over time, TSE’s
program started less than two years ago. This physical industrial automation reduced FTEs and improved
automation technology — which has operators safety and process stability, reportedly leading to
overseeing processes but computer-controlled higher-quality products; AA has achieved similar
machines performing them — has already provided types of stability gains, though without direct labor
major labor efficiency gains: As noted earlier, twenty- reductions to-date.
five years ago, the IJmuiden plant produced 3 million
tons of steel annually with 23,000 FTEs. Today, it Another difference between the legacy systems and
produces 7 million tons of steel with only 9,000 FTEs. AA is the development lead-time: Physical automation
in the steel industry has often been developed
At TSE, the legacy physical automation programs incrementally over many years, while AA can be
address different types of problems than AA and implemented more quickly and requires lower levels
thus provide different benefits: The legacy programs of upfront investment (See Exhibit 7). AA use cases
mostly address highly repeatable tasks, such as have been able to be developed and implemented
moving materials from one area of a factory to in 10 to 12 weeks if there is value potential, relatively
another. This is in contrast to today’s AA use cases, straightforward implementation (e.g., implementing a
which tend to focus on more complex problems that recipe change), and available data to feed the model.26
have traditionally been solved through employees’
experience (in combination with other technology

25
Many have said that broader adoption and advancement of AI could lead to profound workforce changes, suggesting that productivity
gains from AI-related technologies would manifest themselves through labor reductions.
26
Some AA models may take longer to implement if they are directly integrated to control a production process (e.g., an early warning model
on a blast furnace).

30 Observations
Today, the Advanced Analytics use cases offer a example, with no incremental physical automation,
significantly higher return on investment than physical a plant may require an operator to run a train
automation cases at Tata Steel Europe. Because the transporting materials from one location in the plant
IJmuiden plant has been automating its manufacturing to another. With five production shifts per day, this
processes over the past two decades, a large amount would require 5 FTEs total to operate the train. The
of the physical automation potential has already been upfront investment required for physical automation
achieved. What remains for physical automation are significantly outweighs the cost saving benefits
processes that are more difficult to automate, require (reducing workforce by 5 FTEs) (see Call-Out: Physical
heavy investment upfront, and have lower incremental automation example: Torpedo Tapping).
savings opportunities, mostly on labor costs. AA, on
the other hand, addresses different types of problems,
typically not focused on labor reductions but rather
on raw material savings or yield improvement. For

Physical automation example: Torpedo Tapping

A physical automation case at the IJmuiden plant illustrates the difference between the
objectives and nature of physical automation and Advanced Analytics at Tata Steel Europe.

TSE worked for 10 years in total to fully automate a process that required 10 FTEs to operate.
This process involved moving a vessel of liquid hot metal, or a “torpedo car,” from one part of
the plant to a downstream site. Initially, an operator would control the process remotely from
a control room, yet this process was slowly replaced by physical automation, using cranes,
cameras, sensors, and rules-based computer software.

The effort to incrementally automate the process began in 2005. At first, two operators would
control the movement of torpedo cars and the pouring of hot metal from a control room
using joysticks, spread over multiple shifts per day (10 FTE total). By 2015, the process was
fully automated. A fail-safe mode now allows for a human to intervene and take control if the
process fails.

The company faced several challenges in automating this process, including making sure that
the operators’ skills did not atrophy through continued employee training, so they could step in
if the system failed.

In this case, the physical automation led to consistent execution, improved safety, and fewer
incidents and maintenance costs. The incremental value of implementing AA for this use case is
not clear, so there are no plans to use AA-developed models within this process.

31 Observations
5.2.2. Workforce size

Automation technologies have been TSE’s major production) at the same levels of staffing, or more
driver of productivity improvements (including profitable production with similar staffing (e.g., non-
workforce reduction) over the past decades; AA labor cost savings per unit of output).
initiatives, in the past two years, have focused on
identifying next-level potential process improvements Aside from direct labor reductions, companies may
and do not directly focus on labor reduction. So far choose to reap such labor productivity gains through
there have been no known workforce reductions reduced rate of hiring in the long-term future. For
attributable to the AA program at TSE. Instead, as TSE, this consideration holds particular importance,
stated above, the focus has been on product quality since they may face a large voluntary attrition of
improvements, raw material cost savings, and partial their workforce as their operators increasingly reach
process improvements rather than full automation and retirement age.
labor reduction. These areas of focus are expected to
continue, at least in the short-term.

While AA is not expected to result in any layoffs in the

short-term, the cumulative impact of AA initiatives
includes increased labor productivity, in the form of
either or both greater output (e.g., volume of steel

5.2.3. Shifts in workforce tasks

AA use cases are changing the nature of work at

TSE, even when there is no impact on the size of the “Gathering data for the first analysis is far easier
workforce. For example, operators at TSE are typically and far quicker than we used to do it. It used
responsible for several tasks, of which AA may only to take 2-3 months to do a first analysis. At the
be “replacing” one. Therefore, replacement of a task moment, it’s just a click of a button for the first
through AA may free up time for workers to focus on analysis — we can do it in 30 minutes. That’s very
other activities that remain on their docket. In other powerful, and it accelerates the whole process.”
processes in which AA use cases were implemented,
operators’ work appears to have shifted toward Technical Director of the AA program
leveraging recommendations from AA user interface
tools in their activities, instead of relying only upon
their domain expertise as they had previously. In
aggregate, workforce reductions may arise in the long- What do the [Advanced Analytics] teams do with
term future if several more advanced AA use cases excess time [now that data has become more
are centralized and woven together, though in the available and standardized]? First, there is no
interim, it is likely that those plant operators would be shortage of problems to solve and implement.
relocated to different processes in the plant. The focus has shifted to effectively implementing
the analytics versus developing valuable analytics.
For TSE’s data science and engineering teams in As for the mid-to-long term: ‘Too early days to
particular, work has shifted more than for plant comment on that.’
operators - with a change from data collection and
cleaning to data analysis. Technical Director of the AA program

32 Observations
5.2.4. Workforce training

Instead of hiring external talent for its AA program, felt that if it is not understood by senior management,
TSE invested in training about 200 FTEs at various it won’t be successful,” one TSE executive said.
levels of seniority, including the CEO, managers, Advanced data scientists received two extra days of
engineers, IT staff, and staff from non-technical training to allow them to build more complex models.
functions. The company looked internally for talent
rather than externally for several reasons: For one, TSE’s production crews did not receive formal
TSE determined that training internal workers is training but did receive on-the-job training on
more cost-effective and easier than hiring in-demand how to implement recommendations from an AA
scientists from top technology firms. Additionally, model. The AA teams have also started developing
TSE’s employees are proud to work there and were user experience and visualization tools so that the
motivated to “work on the future” of the company, production crew is able to understand the AA models
according to management’s internal surveys on the AA without much data science training.
program. Importantly as well, TSE found that internal
domain experts who had received data science Of note, employees’ wages also did not change
training were much better positioned to develop AA through this process or as a result of the training.
models than outsiders with no steel-sector expertise. According to management, the program was
enthusiastically received by employees: “People are
“We started with training our senior management, way more motivated and they feel they own the work,”
literally with the CEO and then all the way down. We a TSE executive said.

5.2.5. Changes to hiring dynamics

The AA program has also enabled TSE to attract new talent where necessary — such as people with hybrid
profiles in physics and data science. A secondary effect of the initiative is that TSE has developed better
connections to universities, and has now developed a pipeline of new workers through the AA program for
future hires.

5.2.6. Potential medium- to long-term impacts on workforce

While no workforce reduction can be ascribed to the analytics and the physical automation technologies
AA program yet, the medium- to long-term impact mature. Even with this maturing, however, TSE remains
on the workforce from AA at Tata Steel Europe is constrained by contextual factors, such as labor unions
unknown. However, trends seen today may offer or regulations that may influence decisions on whether
some clues. The production workforce is aging, and to pursue labor reduction.
many will retire in the next five to fifteen years, taking
with them experience and knowledge that would be Though it is difficult to estimate which types of workers
difficult to replace through training. This, coupled with are most likely to be impacted at TSE, one may begin
productivity gains from AA, might lead to a slower to speculate. For instance, more-autonomous AA use
hiring rate. cases may replace certain tasks of a plant operators in
the long-term (e.g., early warning model preventing
Some within the company believe that there will likely overboiling) or may require fewer of these operators.
be some labor impact in the coming years as both the On the other end of the spectrum, demand for data

33 Observations
scientists and data engineers will likely increase as 1. Increased demand for data scientists, data
the AA program expands; however, these could be engineers, and business “translators” as the
internal transfers rather than external hires, as in TSE’s Advanced Analytics Program expands to
experiences to-date. Additionally, more-advanced other corporate functions (e.g., HR, finance,
tools and best practices for data engineering procurement).
may entail a need for fewer of these roles as well;
2. Increased demand for hybrid skill sets (e.g.,
extrapolating to future employment figures remains a
backgrounds in both physics and data science).
challenge from TSE’s experiences.
3. Increased need for retraining (e.g., process
Outside of TSE’s steel manufacturing processes, the engineers retrained as data scientists for the AA
company may also see impacts in other domains: For program).
instance, the scope of AA use cases could span other
4. Potential reduction of the production workforce
corporate functions such as HR, procurement, and
in the mid- to long term as AA and automation
finance, requiring shifts in tasks and skills. “I don’t see
programs mature and start to converge; likely no
a single group that will not be affected,” said the AA
workforce reduction in production crew in the short-
program’s technical director.
term.

For now, these micro-trends observed in the TSE 5. Decreasing size of the workforce in the long-term
workforce bear watching: through worker retirements, coupled with lower
hiring rates.

34 Observations
06 Reflections and implications

35 Reflections and implications

6.1. Reflections
6.1.1. Change management

Change management was one of the biggest factors Introducing more advanced analytical techniques
in the successful AA implementation at TSE. The and process automation can naturally introduce fear
management team used several key strategies for or tensions in the workplace. As a project manager
getting employees’ buy-in on the program: reflects, “there have always been people who fear
machines will take over.” On one hand, some within
• Executive sponsorship and clearly defined goals: TSE claim that AA introduces more empowerment
From the program’s inception, Tata Steel Europe and excitement for the production crew. On the other
had multiple executive sponsors driving the change hand, generating trust in new technologies is often
from top to bottom. Management reported that this a challenge: As one interviewee reported, “there are
approach was critical to create the organizational always people that say, ‘I’m an expert in that, are you
momentum, to allocate proper funding and going to replace my expertise?’”
resources, and to develop clear and actionable
goals. The ease of getting employee buy-in also Underlying many of these transitions is a thread of
depended on the desired goals. For example, it communication, which is important in any workplace
was easier to get production crew support for a use environment. TSE found that explanation of the
case focused on raw material savings or overboiling technology, trust in new systems, and transparency
than for one where anonymized data of individual around objectives are important to get buy-in across
workers would have to be used as inputs. different stakeholders. Involving different groups in
the decision-making and development of tools can also
• Cross-functional teams and early involvement:
help to create support and feelings of ownership in the
Having domain experts participate in creating the
transitioning workplace.
AA models led both to better working models and
to increased buy-in from the manufacturing staff.
• Across-the-board training: TSE launched a training
program for 200 of its employees at all levels of
seniority to provide analytical, implementation, and
communication skills and trainings.
• Explainability of models: Especially in the initial
phase of the AA program, operators were more
inclined to implement recommendations that
came from models they could clearly understand.
The technical implementation teams realized that
explainability is a driver of adoption, so they would
often favor less-advanced techniques in models if
the models could then be explained to operators
more easily.

36 Reflections and implications

6.1.2. Business and productivity impact

Since the launch of the Advanced Analytics Program AA is not a panacea — it is a particular set of analytics
at the IJmuiden plant, Tata Steel Europe management tools suited for specific types of problems. At TSE, the
has viewed analytics and machine learning as another project team found that it is best-suited for complex
tool for continuous improvement and a way to stay problems where the optimal solution is still largely
competitive in a highly commoditized and capital- unknown. Analytics models helped “connect the
intensive industry. dots” in processes that were not connected in the
past, pulling together fragmented sources of data and
TSE finds the AA program to have increased produc- standardizing data quality across the organization.
tivity at TSE overall, a key metric in the steel industry, These solutions would help to reduce the reliance
by delivering roughly 13 percent higher EBITDA with upon operators’ intuition, developed through years of
the same production staffing levels. TSE reports that experience.
this has generated significant financial benefits: In its
first year, it generated $5 in financial benefits for every For example, to predict overboiling in the steel plant,
dollar invested. The benefits came largely from savings domain experts would typically have looked for certain
on raw materials and yield improvement. The compa- sounds or furnace characteristics that could signal that
ny also saw indirect margin improvement by shifting an overboiling was imminent; analytics would take
toward higher-quality products, and expects better hundreds of variables, including plant audio data,
customer satisfaction as a consequence. To implement to predict if overboiling would occur. Yet this shift
this program, TSE needed to invest in areas such as does not entirely displace the importance of domain
back-filling labor positions pulled in to deliver the proj- expertise: Instead, the domain experts’ knowledge is
ect; TSE’s IT infrastructure and system upgrades; and in still needed, both to create the initial AA models and
employee training. to check and understand the data.

6.1.3. Workforce impact

While no workforce reductions can be ascribed to the While AA is not expected to result in any layoffs in
AA program today, the potential for medium- to long- the short-term, there is a chance that the production
term workforce reductions could be sizable as the AA workforce could be impacted as Advanced Analytics
initiatives mature and evolve across the organization, use cases mature. In addition, the cumulative impact
as well as the industry. of AA initiatives over the mid- to long term could
potentially result in a reduced rate of hiring. This
In the past, workforce reductions at Tata Steel Europe is especially important for TSE, as it faces a natural
have been achieved through physical automation, reduction in its workforce from upcoming retirements
replacing manual tasks of operators. Yet, as we have and voluntary attrition. The industry’s unionization
learned, different than past physical automation, AA and regulated nature may also directly or indirectly
use cases tend to focus on raw material optimization influence decisions made around AI-related efforts.
or quality improvements, which do not lead to labor
reduction.

37 Reflections and implications

6.2. Conclusion

The steel industry has continuously transformed over the program’s first year. Additionally, TSE was able to
the past decades. Intense competition has pushed make an impact with advanced analytics algorithms
companies to achieve higher operational productivity without having the most sophisticated technology
and efficiency levels, creating leaner organizations and infrastructure or tapping into the external labor market
lowering their cost base. With these goals in mind, for top data science talent. Instead, the company was
steel manufacturers have redesigned their processes, able to train its existing technical teams and equip
automated manual tasks to the extent possible, and them with the skills needed.
made their operations more efficient. As a result,
the industry has seen an ongoing decline in direct Yet TSE’s Advanced Analytics integration was not
employment over the past half century. In the U.S., for without challenges; issues such as the availability,
example, labor productivity (output per worker) grew connectivity, and quality of data; systems and
by a factor of five from 1962 to 2005, which enabled IT readiness; organizational buy-in; and change
significant workforce reductions. management needed to be overcome to help steward
this business-value.
Continuous improvement and innovation have been
Tata Steel Europe’s guiding principles for many years For labor specifically, the case of Tata Steel Europe
as a result of operating in this commoditized, highly- raises questions about the future of the workforce
competitive, and capital-intensive industry. Over the in the steel industry (See Appendix 7.1 for open
past decades, the company undertook major physical questions for further research). Though the AA
automation and operational efficiency programs. In implementations did not lead to any layoffs, this was
the past two years, it identified Advanced Analytics largely due to the unionization of the workforce at
applications and use cases to continue this ongoing TSE. Yet, the company was still able to increase worker
improvement, though with somewhat different productivity through improved yield and more efficient
features than past transformations. use of raw materials. In the future, these improvements
may lead to workforce reductions, perhaps through
TSE’s use of Advanced Analytics demonstrates how retirements and voluntary attrition as TSE’s workforce
AI can create value in capital-intensive industries, ages. While TSE illustrates a positive use case of
such as in the steel industry. After investing in its a company that was able to successfully re-skill its
IT infrastructure, compiling its data sources, and own workforce internally and introduce AI without
standardizing its data quality, TSE reports that it was any layoffs, it remains to be seen whether this will
able to quickly implement key Advanced Analytics use be replicable at scale and in contexts without strong
cases and reach ROI levels of $5 for every $1 spent in union protections.

38 Reflections and implications

07 Appendix

39 Appendix Introduction
 7.1. Questions for future research

The Tata Steel Europe case study raises a number of questions that could be
addressed in future research.

One question of note is the likely progression for labor as Advanced Analytics use
cases become more mature: For instance, in one system, what began as a predictive
early-warning tool for operators’ use has evolved to a fully autonomous system, in
which an operator does not intervene. As the AI-related technologies at TSE mature,
it is likely that they will start to intersect with existing automation programs or
expand to further use cases outside of production (e.g., use cases in procurement,
finance, HR), which may lead to job loss, or may allow employees to focus on other
areas of the company’s operations.

Complicating these dynamics is TSE’s aging workforce: If TSE does not reduce its
dependence on labor, how might the company be affected in coming years by lower
availability of workers? Will there be further economic savings in the form of raw
materials and yield, or will the AI wave look similar to the physical automation wave
of the past few decades, leading to reductions in the workforce? It is unclear what
these questions will mean in sum for the future of work at TSE’s IJmuiden plant and
for incumbents of older industries in general as they adopt AI to remain competitive.

40 Appendix
7.2. Exhibit 2. Business summary of the IJmuiden plant
Exhibit 2: Business summary of the IJmuiden plant for Tata Steel

Source: Tata Steel Europe, Sustainability Report, 2017

Source: Aerial image of the IJmuiden plant in the Netherlands (Source: Tata Steel Europe)

41 Appendix
7.3. Exhibit 5. Global steel production
Exhibit 5: Global crude steel production, 1950 to 2015

SOURCE: World Steel Dynamics (Feb. 2015)

Exhibit 5B: Direct employment of the steel industry is declining

SOURCE: EUROFER, State of Steel (2018)

42 Appendix
7.4. Exhibit 6. Macroeconomic context of the steel
industry in Europe
Exhibit 6A: The steel industry is one of the largest contributors to EU manufacturing

Source: Eurofer (2017)

Exhibit 6B: EU steel industry direct economic impact

1 Values relate to “manufacture of basic iron and steel and

of ferro-alloys” (NACE 24.1)
2 Gross spending on machinery, equipment and
construction work only. Excludes land, existing buildings,
and intangible assets

Source: eurofer (employment data); Oxford Economics estimates based mainly on Eurostaes’s “Annual detailed enterprise statistics for
industry”Eurostat, EUROFER; Oxford Economics Sectors defined by NACE rev 2 code in brackets

43 Appendix