How AI Is Reshaping
Five Manufacturing
Industries
ByAyesha Khanna
, Former Forbes Councils Member.
for Forbes Technology Council
COUNCIL POST | Membership (fee-based)
Jan 17, 2024, 07:00am EST
Dr. Ayesha Khanna is the CEO of Addo.
Today, digital transformation is a cornerstone of survival for
businesses.
From automated factories to AI quality control, the primary
objective of digital transformation is forging a competitive edge
through technology, resulting in enhanced customer experiences
and reduced operational costs.
The manufacturing industry is at the forefront of digital
transformation, leveraging technologies like big data analytics, AI
and robotics. The results are tangible, according to McKinsey, who
found that machine downtime can be reduced by 30% to 50% and
quality-related costs can be reduced by 10% to 20%, among other
benefits.
In this article, I'll explore how five industries use AI in
manufacturing, and what manufacturing leaders need to know
about what's next for the industry.
Automotive Industry
Automotive manufacturing requires precision and accuracy, and AI
can help enhance that.
For example, Ford employs cobots for welding, gluing and quality
control tasks. It uses six cobots to sand the entire body surface of a
car in 35 seconds. Similarly, BMW's Spartanburg plant, producing
60% of U.S. BMWs, uses AI-managed robots, saving $1 million
yearly and reallocating workers.
The automotive AI market is projected to hit $7 billion by 2027,
highlighting it as one of the leading industries in adopting AI in
manufacturing.
Electronics Industry
Electronic manufacturing also requires precision due to its intricate
components, and AI can be critical in minimizing production errors,
improving product design and accelerating time-to-market.More
For instance, Samsung's South Korea plant uses automated
vehicles (AGVs), robots and mechanical arms for tasks like
assembly, material transport, and quality checks for phones like
Galaxy S23 and Z Flip 5. These tools can help companies maintain
high-quality standards, including inspections of 30,000 to 50,000
components.
Nvidia is using AI to optimize the placement of intricate transistor
configurations on silicon substrates, which not only saves time but
offers greater control over price and speed. It proved its efficiency
by optimizing a design featuring 2.7 million cells and 320 macros in
just three hours.
With a vast market and continued AI innovation, enhanced use of AI
involvement is becoming table stakes for companies manufacturing
electronics.
Aerospace And Defense Industry
AI-driven manufacturing enhances product safety and reliability by
producing precise components, boosting performance and system
safety. The AI in aviation market was worth $686.4 million in 2022
and is expected to grow at a CAGR of over 20%.
Airbus, with Neural Concept's tech, cut aircraft aerodynamics
prediction time from one hour to 30 milliseconds using ML. This
kind of productivity boost can enable design teams to explore
10,000 more changes in the same time frame as the traditional
computer-aided engineering approach.
Likewise, Rolls-Royce, in collaboration with IFS, uses AI in
aerospace manufacturing through the Blue Data Thread strategy.
This approach utilizes digital twins and AI for predictive
maintenance, resulting in a 48% increase in time before the first
engine removal.
Food And Beverage Industry
Food and beverage production requires advanced quality
assurance, particularly in the fast-moving consumer goods (FMCG)
sector, due to its “high-speed” nature. Equipment breakdowns and
faulty products can hinder that; however, integrating AI can boost
efficiency, cost-effectiveness and product quality and safety.
The global AI market for the food and beverage industry is set to
reach $35.42 billion by 2028.
Startups specializing in predictive maintenance technology are
particularly in demand. Take Augury Inc., for instance. They helped
PepsiCo’s Frito-Lay gain 4,000 hours of manufacturing
capacity annually through its predictive maintenance systems that
decreased unplanned downtime and costs at four Frito-Lay plants.
Pharmaceutical Industry
It usually takes a decade to develop a drug, plus two more years for
it to reach the market. Unfortunately, 90% of drugs fail in the
clinical trial phases, resetting the clock. AI can accelerate drug
development and enhance quality control.
Pfizer, for instance, using IBM's supercomputing and AI, designed
the Covid-19 drug Paxlovid in four months, reducing computational
time by 80% to 90%.
Here are three of the areas where AI can alleviate drug-discovery
challenges:
1. Protein Structure Prediction: AI systems like AlphaFold2 have
transformed protein structure prediction, enabling researchers to
understand the blueprints of complex molecules accurately and
potentially saving years of lab work.
2. Function Forecasting: A recent Forbes article explored how AI
models can predict how large molecules function and understand
how proteins bind to their targets and the movement of antibodies,
enhancing the development of therapeutic responses.
3. New Therapies Design: AI algorithms use vast data to design
proteins, antibodies and mRNA structures for diseases like cancer.
Genesis Therapeutics, for example, uses AI to design and predict
the effectiveness, specificity and potential side effects of new
medication.
AI in drug development could generate 50 new drugs, resulting in
$50 billion in sales over a decade. Over 80 companies are
advancing AI-driven drug development, attracting investments from
pharmaceutical giants.
The Way Forward
A recent survey conducted by Augury of 500 firms reveals that 63%
plan to boost AI spending in manufacturing. This aligns with AI in
manufacturing market projection, which is estimated to
reach $20.8 billion by 2028, according to MarketsandMarkets.
The efficiency gains from AI integration translate into cost and time
savings, allowing resources to be redirected to more critical tasks
and opportunities.
AI in manufacturing relies on three pillars: problem, people and
process. Here are five steps that can help ensure that you develop
all three pillars:
1. Define the problem: Identify inaccuracies causing higher costs.
New-to-AI companies should break down the issue into SMART
goals and assess AI's potential for long-term cost savings.
2. Address resources and data: Form a diverse team of technical
and business experts. Evaluate in-house capabilities and consider
outsourcing or hiring. Verify data sufficiency, clean and structure
the data and decide on storage solutions.
3. Assess data quality: Is the data modern, accessible and
sufficient? Modify as needed.
4. AI model considerations: Decide to build, buy off the shelf or
adopt a hybrid approach.
5. Fine-tuning and deployment: Discuss model refinement,
deployment and scalability.
As a final note, be sure to adhere to ethical guidelines and
frameworks throughout each step, as it is crucial to rigorously
check for bias and develop guardrails against it.
