AI in Manufacturing: Opportunities & Challenges

AI IN MANUFACTURING:
OPPORTUNITIES &
CHALLENGES
TATHAGATVARMA
VP STRATEGY, WALMART GLOBALTECH | PHD SCHOLAR (EFPM), INDIAN SCHOOL OF BUSINESS (ISB)

DISCLAIMER THESE ARE MY
PERSONALVIEWS!

PROJECTED
ECONOMIC
IMPACT OF
$15 TRILLION
BY 2030…!!!
http://preview.thenewsmarket.com/Previews/PWC/DocumentAssets/476830.pdf

BUSINESS
BENEFITS OF
AI
https://hbr.org/2018/01/artificial-intelligence-for-the-real-world

VIRTUALLY ALL
ASPECTS OF
MOST
BUSINESSES…
https://info.algorithmia.com/2021

AI CAPABILITIES
IN BUSINESS
PROCESSES

BUT,THERE’S ALSO A
LOT OF HYPE!!!
https://www.gartner.com/en/articles/what-s-new-in-artificial-intelligence-from-the-2022-gartner-hype-cycle

…WITH ALARMINGLY HIGH FAILURE RATES!
• Gartner: 85% of AI projects will fail and deliver erroneous outcomes through 2022.
• MIT SMR: 70% of companies report minimal or no impact from AI.Among the 90%
of companies that have made some investment in AI, fewer than 2 out of 5 report
business gains from AI in the past three years.
• VentureBeat: 87% of data science projects never make it into production.Tom
Siebel says 99% of internal AI projects fail!
• Andrew Ng (HBR): It is not unusual for teams to celebrate a successful proof of
concept, only to realize that they still have another 12-24 months of work before the
system can be deployed and maintained.
https://research.aimultiple.com/ai-fail/

…AND LOW ROI & LONG PAYBACK PERIODS!
The ROI for AI projects varies greatly,
based on how much experience an
organization has.
Leaders showed an average of a 4.3%
ROI for their projects, compared to only
0.2% for beginning companies.
Payback periods also varied, with leaders
reporting a typical payback period of
1.2 years and beginners at 1.6 years.
https://www2.deloitte.com/us/en/insights/industry/technology/artificial-intelligence-roi.html

SO,WHAT IS AI?
HINT:THERE IS NOTHING REALLY “INTELLIGENT” ABOUT IT!

WHAT IS “ARTIFICIAL INTELLIGENCE”, OR AI?
Prof John McCarthy, Father of Artificial Intelligence:
• Intelligence is the computational part of the ability to achieve goals in the
world.Varying kinds and degrees of intelligence occur in people, many animals
and some machines.
• Artificial Intelligence is the science and engineering of making intelligent
machines, especially intelligent computer programs. It is related to the similar
task of using computers to understand human intelligence, but AI does not have
to confine itself to methods that are biologically observable.
http://jmc.stanford.edu/artificial-intelligence/what-is-ai/index.html

GOOD OLD FASHIONED AI (GOFAI)
• Symbolic artificial intelligence is the term for the collection of all methods in artificial intelligence research
that are based on high-level symbolic (human-readable) representations of problems, logic and search. Symbolic AI
was the dominant paradigm of AI research from the mid-1950s until the late 1980s.[1][2]
• John Haugeland gave the name GOFAI ("Good Old-Fashioned Artificial Intelligence") to symbolic AI in his 1985
book Artificial Intelligence:TheVery Idea, which explored the philosophical implications of artificial intelligence
research. In robotics the analogous term is GOFR ("Good Old-Fashioned Robotics").[3]
• Researchers in the 1960s and the 1970s were convinced that symbolic approaches would eventually succeed in
creating a machine with artificial general intelligence and considered this the goal of their field.
• However, the symbolic approach would eventually be abandoned, largely because of the technical limits of this
approach. It was succeeded by highly mathematical Statistical AI which is largely directed at specific problems with
specific goals, rather than general intelligence. Research into general intelligence is now studied in the exploratory
sub-field of artificial general intelligence.
Symbolic Artificial Intelligence, https://en.wikipedia.org/wiki/Symbolic_artificial_intelligence

AI EVOLUTION TIMELINE…
https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/

KEY MILESTONES
• 1943: Walter Pitts & Warren McCullogh develop a computer model based on Neural Networks of human brain using a combination of algorithms and maths they called “threshold logic” to mimic the
thought process.
• 1950: Alan Turing proposes the imitation game, aka “Turing Test”
• 1952: Hodekin-Huxley paper of brain as neurons forming an electrical network
• 1956: John McCarthy coins the term “Artificial Intelligence” and organizes Dartmouth Summer Research Project, the first conference on AI.
• 1960s: Research labs established at MIT, Stanford, SRI, etc. to mimic human intelligence by problem-solving or playing games like checkers or chess.
• 1960: Henry Kelley develops the basics of continuous Back Propagation (backprop) model
• 1962: Stuart Dreyfus develops chain rule to simplify backprop.
• 1965: Alexey Grigoryevich Ivakhnenko & Valentin Grigorʹevich Lapa develop Deep Learning algorithms using polynomial activation functions and statistical analysis at each layer.
• 1970s: MYCIN was able to diagnose certain kinds of bacterial infections based on symptoms input.
• 1970s: A “prospector” expert system uncovers a hidden mineral deposit of porphyr molybdenum (a form of copper deposit) at Mount Tolman in the state of Washington.
• 1973-80s: First “AI Winter”
• 1981: John Searle proposes “Chinese Room”
• 1980s: Development of Expert Systems bring some successes (e.g. DEC’s XCON)
• 1985-90s: Second “AI Winter”
• 1979: Kunihiko Fukushima develops “Neocognitron” the first Convolutional Neural Network (CNN) with multiple pooling and convolutional networks that allows computer to “learn” to recognize visual
patterns using manually-adjustable “weights” of certain connections.
• 1990s: Focus shifts to “Intelligent Agents”
• 1997: IBM Deep Blue beats World Chess Champion Garry Kasparov (Artificial Intelligence)
• 2011: IBM Watson beats human players on US game show Jeopardy (Machine Learning)
• 2012: Deep Learning
• 2014: Ian Goodfellow creates Generative Adversarial Networks (GANs)
• 2016: Google’s AlphaGo beats boardgame Go master Lee Sedol (Deep Learning)

EVOLUTION OF AI, ML, DL
https://www.stateofai2019.com/summary

AI, ML, NN, DL…
• Machine learning, deep learning, and neural networks are all sub-fields of artificial
intelligence. However, deep learning is actually a sub-field of machine learning, and neural
networks is a sub-field of deep learning.
• The way in which deep learning and machine learning differ is in how each algorithm learns.
Deep learning automates much of the feature extraction piece of the process, eliminating
some of the manual human intervention required and enabling the use of larger data sets.
• Classical, or "non-deep", machine learning is more dependent on human intervention to
learn. Human experts determine the set of features to understand the differences between
data inputs, usually requiring more structured data to learn.
Machine Learning, https://www.ibm.com/cloud/learn/machine-learning

MACHINE LEARNING
• Term coined by Arthur Samuel in 1959
• Machine learning is a branch of artificial intelligence (AI) and computer
science which focuses on the use of data and algorithms to imitate the
way that humans learn, gradually improving its accuracy.
• Through the use of statistical methods, algorithms are trained to make
classifications or predictions, uncovering key insights within data mining
projects.These insights subsequently drive decision making within
applications and businesses, ideally impacting key growth metrics.
https://www.ibm.com/cloud/learn/machine-learning

POTENTIAL OF AI
•AI is the new electricity – Andrew Ng
•AI is more profound than fire or electricity –
Sundar Pichai
•People should stop training radiologists – Geoff
Hinton, 2016, https://youtu.be/2HMPRXstSvQ

WHERE TO USE AI?
If a typical person can do a mental task with
less than one second of thought, we can
probably automate it using AI either now or
in the near future. – Andrew Ng
https://hbr.org/2016/11/what-artificial-intelligence-can-and-cant-do-right-now

SO,WHY NOW?
After decades of start/stops, finally AI seems to be at the cusp of its (third) resurgence.

AI IS THE NEW SOFTWARE!
• Machine learning was a huge leap from programmed instructions and if-then statements
that merely simulated the very human process of thinking and making decisions.
• With machine learning, the machine no longer needs to be explicitly programmed
to complete a task; it can pour through massive data sets and create its own
understanding. It can learn from the data and create its own model, one that represents
the different rules to explain relationships among data and use those rules to draw
conclusions and make decisions and predictions.
• A machine learning algorithm is a mathematical function that enables the machine to
identify relationships among inputs and outputs.The programmer’s role has shifted
from one of writing explicit instructions to creating and choosing the right
algorithms.
Rose, Doug.Artificial Intelligence for Business:WhatYou Need to Know about Machine Learning and Neural Networks.

HOW IS AI DIFFERENT FROM SOFTWARE?
Traditional Software AI Software
Reasoning Deductive Inductive
Inputs Data + Program Data + Output
Logic Manually pre-programmed to perform a specific task on
a given dataset
Programmed to automatically keep learning rules from a
given dataset
Output Output Models, Rules
Learning Learns one-time from the programmer “Learns” constantly being the data
Resource Code Data
Solutions Deterministic Probabilistic
Output Consistently remains the same Can improve with usage (or degrade over time)
Business model One-time development efforts, followed by multiple
sales, and small maintenance effort (optional)
Each project is one-off, and needs full lifecycle
management mandatorily

SUPERVISEDVS UNSUPERVISED LEARNING

WHERE TO USE AI?
• If a typical person can do a mental task with less than one
second of thought, we can probably automate it using AI either
now or in the near future. – Andrew Ng,
https://hbr.org/2016/11/what-artificial-intelligence-can-and-cant-
do-right-now

LIMITATIONS OF AI
When people talk about AI, machine learning, automation, big data, cognitive computing, or
deep learning, they’re talking about the ability of machines to learn to fulfill objectives based
on data and reasoning.This is tremendously important, and is already changing business in
practically every industry. In spite of all the bold claims, there remain several core problems at
the heart of Artificial Intelligence where little progress has been made (including learning by
analogy, and natural language understanding). Machine learning isn’t magic, and the truth is we
have neither the data nor the understanding necessary to build machines that make routine
decisions as well as human beings. https://hbr.org/2016/11/how-to-make-your-company-
machine-learning-ready

FAILURE MODES OF AI
• Brittleness
• Embedded Bias
• Catastrophic Forgetting
• Explainability
• Quantifying Uncertainty
• Common Sense
• Math
Source: https://spectrum.ieee.org/ai-failures

CHALLENGES
• Unavailability of Skills &Talent
• Technology is quite good, but still maturing
• Data is often siloed or not available, or poor quality, etc.
• Business case, and alignment of strategy and business model to an AI-firm
• Societal concerns like ethics, privacy, transparency, bias, surveillance, etc.
• Governance issues such as data management, regulations, legal frameworks, etc.
• …..and many many more…!!!

AI IN MANUFACTURING
• The artificial intelligence (AI) revolution allows the conversion of large amounts of
data into actionable insights and predictions that can provide impetus to data-driven
processes.
• Manufacturing companies capture value from AI using different mechanisms, the
most common being eliminating redundant work, solving existing problems and
revealing hidden value by analysing and recognizing patterns in data.
• AI is applied to augment tasks such as classification, continuous estimation,
clustering, optimization, anomaly detection, rankings, recommendations and data
generation to solve industrial problems.
Unlocking value from Artificial Intelligence in Manufacturing,World Economic Forum, Dec 2022

AI IN MANUFACTURING CONTEXT
• In principle, AI could unlock more than $13 trillion in the global economy and
boost GDP by 2% per year. However, companies struggle to tap into the value
that AI applications can create.
• Even though the impact of AI applications on manufacturing processes is known,
the full opportunity from their deployment is still to be uncovered due to a number
of organizational and technical roadblocks.

VALUE CREATION USING AI IN MANUFACTURING
• Operational Performance
• Workforce Augmentation
• Sustainability

OPERATIONAL PERFORMANCE
By automating and optimizing routine
processes and tasks, increasing
productivity and operational efficiencies,
improving quality (e.g. reducing defects,
forecasting unwanted failures) and
optimizing production parameters

WORKFORCE AUGMENTATION
By guiding the decision-making process
and parameter setting, enhancing the
accuracy of predictions and forecasting,
reducing repetitive tasks and increasing
human-robot interactions

SUSTAINABILITY
By optimizing material and energy usage,
increasing energy efficiencies, reducing
scrap rates and extending machine
lifespans.

SAMPLE USE
CASES AND
IMPACT
• Time efficiency: Alarm
rationalization meetings shortened
from 4 hours to 30 minutes
• Unsafe situations and actions
reduced by 70-80%
• Costs reduced by 60% by
preventing the use of excess
welding materials
• Productivity increased by 30-40%
• Reduction of false-reject rate by
an average of 88%
• Number of quality stops reduced
from 81 to 20 per week
• Downtime reduced by 90%

WHERE’S THE BUSINESSVALUE?
• 70% of respondents understand how AI can generate business value
• 59% have an AI strategy in place
• 57% affirm that their companies are piloting or deploying AI.
• Despite these trends, only 1 in 10 companies believe they generate significant
financial benefits with AI.

CHALLENGES IN ADOPTION OF AI IN
MANUFACTURING

MISMATCH BETWEEN AI CAPABILITIES AND
OPERATIONAL NEEDS
• Manufacturers have often selected AI projects based on existing technical capabilities
instead of focusing on the impact on business operations. The match between
business pain points and AI technologies is not always thoroughly considered.
Therefore, AI solutions may be technically feasible but fail to solve a relevant,
impactful problem in operations. This causes a mismatch of expectations and
hinders their wider adoption in manufacturing.
• Building a solid business case with a problem-oriented approach that clearly defines
business needs and evaluating the value of an AI solution compared to alternative
solutions are the first steps in overcoming that barrier to adoption and scale.

ABSENCE OF A STRATEGIC APPROACH AND
LEADERSHIP COMMUNICATION
• A clear company-wide AI strategy and communication plan are often ignored.
Without the right sponsors and committed leaders to start the dialogue and collect
the buy-in from end-users, the onboarding of AI applications across the
company can’t occur due to workforce reluctance.
• As AI is changing the ways of working, communicating the strategic approach,
benefits and new processes can help increase endusers’ willingness to embrace
it in their routines.

INSUFFICIENT SKILLS AT THE INTERSECTION
OF AI AND OPERATIONS
• External consultants or information technology (IT) experts who have a limited
understanding of the manufacturing requirements on the shop floor often lead AI
projects.
• However, to be successful, AI applications require development and
implementation by cross-functional teams with diverse expertise at the
convergence of IT, operational technology (OT), data and AI technologies. This
requires upskilling the workforce and attracting new talent in manufacturing.

DATA AVAILABILITY AND THE ABSENCE OF A
DATA GOVERNANCE STRUCTURE
• Applying machine learning models requires training on large amounts of data to
recognize patterns and relationships. However, manufacturing companies often
rely on small data sets and fragmented data, hindering the accuracy of the
resulting insights. Even when available, these data sets may not represent
appropriate failure cases or relevant process situations and are mostly not
interoperable. Creating a single source of information ensures that businesses
operate based on standardized, relevant data across the organization.
• To overcome this challenge, sharing data across companies’ boundaries can support
joint efforts to adopt artificial intelligence techniques in the manufacturing sector and
rely, in turn, on a set of organizational and technological success factors.

LACK OF EXPLAINABLE AI MODELS IN
MANUFACTURING
• The perception of AI models as complex, nontransparent and uninterpretable systems
hinders their deployment. Manufacturers need AI models that are either open and
transparent to build trust in the predictions and specific results or interpretable for
domain experts to accept them. AI-provided predictions need to be meaningful,
explainable and accurate and have a warning mechanism in place to minimize
risks.
• Explainable AI tools and techniques allow experts to obtain justifications for their
results in a format that manufacturing users can understand. The greater the
confidence in the AI-powered output, the faster and more widely AI deployment can
happen.

SIGNIFICANT CUSTOMIZATION EFFORTS
ACROSS MANUFACTURING USE CASES
• Factories are complex engineered systems and AI models need configuration to
be adapted to each process and conform to its constraints. Hence, it is not
possible to simply apply trained AI models or pipelines from one
manufacturing use case to another. The design of the machine learning
pipeline and the pre-processing, training and testing of AI models still need
manual intervention for customization, which is not yet fully automated.
• Additionally, industrial companies struggle to find commercially available
hardware and software with off-the-shelf AI features that require minor
customization.

RECAP
• Today’s AI is not really “AI”! At most, it is what we call as “Machine Learning” (ML). But, let’s
stick with AI for now J
• In the broadest sense, it is a “General Purpose Technology” (GPT). Just like fire, electricity or
computers.
• AI has huge potential to solve problems that could benefit from high-speed and at-scale data-
based decision-making for routine and repetitive tasks , especially for “noisy data” conditions.
• Perhaps the first trillion-dollar economy that won’t require natural resources, financial capital,
or human labor.
• Challenges with adoption notwithstanding, there are huge opportunities for businesses to
leverage AI to boost productivity and deliver financial returns.

AI in Manufacturing: Opportunities & Challenges

More Related Content

What's hot (20)

Similar to AI in Manufacturing: Opportunities & Challenges (20)

More from Dr. Tathagat Varma (20)

Recently uploaded (20)

AI in Manufacturing: Opportunities & Challenges