By Yannick Schilly, President and CEO and Ramesh Chandra, Partner Supply Chain
Although it has been exactly 30 years this month since I started my career in the industrial world, and specifically in the industrial automation and digitalization world, November 1994 seems like yesterday! I have since seen industry evolve dramatically and for the best, but lately I have witnessed an incredible acceleration in technological development and technology adoption – nothing quite compares to the pace of adoption of artificial intelligence.
In 1942, Joseph Schumpeter coined the term “creative destruction,” which explains the theory behind the acceleration of innovation cycles as shown in this chart. All those innovations have helped industry to address new challenges, develop better products, solve more technical problems, gain force, speed, power, but also permanently reduce costs and manufacturing and development cycles, while reducing lead-times and time-to-market.
The current iteration of industrial revolution – Industry 4.0 – which is about automation, digitalization and ultra-connected world (Internet of Things), autonomous cyber-physical systems and industrial AI, is estimated to take only between 15 and 20 years for a broad adoption in industry. This is, of course, not the final cycle, with the next one already in preparation and expected to be much shorter than Industry 4.0.
So, a few key questions stressing out manufacturing leaders are:
Does it matter to our company or is this all hype around Chat GTP?
Chat GTP has seen an explosive adoption rate, and if you’re reading this article, the chances are high you have tested it and that your employees are using it extensively. Big tech – Amazon, Nvidia, Apple, Microsoft, Facebook, TESLA, etc. – are all investing massively in AI, and for very good reasons (by the way, mainstream companies like Walmart, Coca-Cola, GE, 3M and P&G do the same). It is about disruption and competitive advantages. Not only for their own businesses, but more importantly for their customers, who are in the race for adopting the best technology to create and sustain competitive advantages.
But what could industrial AI do for us if we are in manufacturing or supply chain?
Industrial AI, as important as “office AI,” will empower employees to gain productivity and reduce costs and lead-times. Its application, quite literally, impacts every corner of the company, wherever data is processed:
- Sales and marketing
- Product design
- Sourcing and supply chain management
- Production and operations
- Quality management
- Logistics
- Finance, reporting and analytics
- Human resources management
Here are some interesting statistics I’ve selected for you from large institutional players:
- PwC reports that AI could contribute upwards of $15.7 trillion to the global economy by 2030, through $6.6 trillion from productivity increase and $9.1 trillion from consumption-side effects.
- Deloitte estimates that manufacturers planning to invest in AI expect to increase their revenues by an average of 11.6% over the next 3 years.
- McKinsey Global Institute analysis suggests that AI adoption within manufacturing is accelerating, with 50% of companies in automotive and electronics (who are typically considered early adopters) already integrating AI technology in production and core business operations.
- Accenture forecasts that AI will increase business productivity by up to 40% by automating processes and optimizing tasks.
- KPMG emphasizes that AI in industrial settings has the potential to reduce equipment downtime by up to 50% through predictive maintenance.
- In pharmaceuticals, BCG predicts that drug development costs will be reduced by 30% through faster and more efficient discovery processes and clinical trials (R&D).
This list could go on, but I’m guessing you see my point about the state of this matter.
So, does AI matter to industrial companies of all sizes? Yes, quite simply because one does not have the choice or luxury to ignore a mega-trend like AI that is all-pervasive.
So, what does AI and Industrial AI really mean – what are the core concepts behind it?
Artificial intelligence, just like its literal meaning, is computer-based intelligence, which like human intelligence, refers to the complex process through which computer-based systems are trained to “learn” and recognize patterns from available data.
We as humans, learn over a lifetime through many ways:
- We learn to crawl, walk and run.
- We learn to socialize and communicate.
- We learn a trade and/or go to college.
- We keep learning through our professional life and through life experiences.
Not all information we consume is beneficial as at times we consume junk information or fake information.
The same way, industrial AI powered by the modern computing power and supported by Moore’s Law (increasing processing power of microchips) could process an incredible amount of information, along with efficient algorithms to process data (which is pre-scrubbed, contextualized and of excellent quality) in ridiculously small amounts of time, to support humans in their decision making.
Data sources are our industrial digital systems like:
- Our office applications.
- Our enterprise systems like ERP, CRM, PLM, HRIS, SCM and other systems.
- Our industrial data (also called IIOT – Industrial Internet of Things), which is collected in real time from machine data out of production equipment and feeds an internal network made of sensors/actuators/HMI’s/PLC’s/SCADA systems and MES Systems (manufacturing Execution Systems), enabling continuous improvement of operations based on data and on machine learning algorithms and artificial intelligence.
- Our product data, whether we produce aircraft engines, cars, machine tools or industrial kitchen equipment.
Here are few examples of industrial AI use cases:
- The use of simulation and digital twins combined with AI backed PLM systems and design libraries increases the efficiency and effectiveness of the innovation process.
- The use of AI in supply chain management allows us to process big data volumes to simulate complex network optimization models and simulate total cost of ownership based on global sourcing and costing algorithms, as well as to improve demand planning, forecasting (S&OP) and inventory management.
- Production gets optimized through tools, enhancing the shopfloor operators ability to minimize waste in looking for information or materials and also accelerates learning cycles and making recommendations on how to operate tasks and remind employees of critical tasks (like your car today reminds you to check the oil level or the tire pressure).
- Quality control through visual inspection and highly automated execution of complex test protocols (imagine a high speed cartoner in the beverage production sector), where production cycles are so fast that no human beings have any chance to compete with technology, nor would they want to do such boring and debilitating tasks.
- Predictive maintenance helps to reduce downtime and save companies billions of $$.
So how to approach this beast that seems complex and highly technical? Where do I start?
The journey is fortunately not impossible, we just have to be systematic and approach it in the right sequence of events:
- Start with a baseline analysis and identify where you stand as a company on a technological maturity level (automation and digitalization), but also from a process and skills level within the organization.
- Identify areas of opportunity delivering tangible results.
- Set up the right internal organization and make it a leadership priority (the C-suite needs to stay involved).
- Identify a clear roadmap with clear milestones for success.
- Allocate resources to the initiatives (people and capital).
- Define small pilots and roll-out after demonstrating success.
- Establish the proper steering platforms and executive sponsors.
- Address a clear vision for the organization and address change management and internal communication to ensure to embark people on a journey for change.
- Celebrating milestones and pilot areas of success.
What are the key barriers to success?
In my view those are a few examples of common barriers:
- Leadership commitment is not fully there.
- Vision and objectives are not clearly defined/understood.
- No proper internal set-up/organization/resource allocation.
- Lack of skills and expertise/no outside coach or expert support.
- No clear strategy, roadmap or plans.
- Lack of holistic approach and cross-functional collaboration/siloed organization.
- Cultural change management is not properly addressed and communication to employees is not clear, creating fear and anxiety.
- Unrealistic expectations are set/lack of strategic patience.
- Lack of proper sequence of evolution (organizational maturity).
- Lack of on-going investments in training and education to keep the workforce up to speed.
- Big gaps in process/data and systems, and unwillingness to address them.
Here as well I could go on with a long list, but all of those risks can be properly addressed in setting up a digital transformation with an AI focus on the right foundations.
Are you ready?
I hope this brief article helps you to reflect on the subject and spark some thoughts.
If you want to learn more about digital transformation and industrial AI, do not hesitate to reach out to plan for a discussion and to request our white paper on the subject.
Wishing you all the best and a successful industrial AI transformation journey!
Yannick Schilly and Ramesh Chandra
