Artificial Intelligence (AI) has rapidly infiltrated various industries, and the manufacturing sector is no exception. The integration of AI is in the process of revolutionizing manufacturing, and helping companies realize greater value through increased efficiency, productivity, and quality.
Progress thus far in AI applications has been hard-fought. As a new technology, we are on the battlefield front lines, working against practicalities of applications and pushing our understanding of what might be possible… and profitable, when it comes to leveraging this new technology.
Indeed, the initial adoption of AI in manufacturing was primarily focused on specific tasks, such as quality control and predictive maintenance. In this way, AI has been with us for a long time, but only in the past couple years has seen any substantial attention in the mainstream media. In the early days, for example, AI-powered vision systems were deployed to inspect products for defects, while machine learning algorithms were used to predict equipment failures. These systems relied on basic sensors and were heavily dependent on detailed programming of the systems that was undertaken for the tasks at hand. However, these early implementations often faced challenges, including data quality issues, lack of skilled personnel, and concerns about the cost of implementation.
Surge in AI Adoption
Over the past decade, the growth of AI in manufacturing has been exponential. The global market for AI in manufacturing is projected to grow significantly, reaching $68 billion by 2032 from $5 billion in 2023 (Precedence Research). Several factors have contributed to this surge, not the least of which is advancement in computing power and the sophistication of AI technology itself. Breakthroughs in machine learning, deep learning, and natural language processing have made AI more powerful and versatile. The other side of the coin of this trend is the decreasing cost of AI hardware and software, making the technology more accessible to manufacturers of all scales.
This acceleration of adoption has only increased as manufacturers have become more aware of the potential benefits of AI and its ability to address key challenges, driven largely by media attention that has been given to the technology. Now we see, more and more, emerging success stories of companies that have implemented AI in their manufacturing processes, which itself has inspired others to follow suit.
Key Applications of AI in Manufacturing
AI is being used in a wide range of manufacturing applications, including predictive maintenance, quality control, supply chain management and optimization, robotics and product development. AI is expected to increase productivity by 40% or more in the manufacturing industry by 2035 (Accenture), and another study by PwC determined that AI-enabled predictive maintenance can reduce maintenance costs by up to 30% and unplanned downtime by 45%.
As for its practical capabilities, AI-driven systems can analyze sensor data from machines to predict potential failures, allowing for proactive maintenance and reducing downtime. Now, more than just limited functionality sensors, AI-powered vision systems can inspect products for defects with greater accuracy and speed than human inspectors.
AI is being applied in processes, production oversight, tracking and is being used to optimize supply chain processes by forecasting demand, managing inventory levels, and improving logistics. AI-enabled robots can perform tasks with greater precision and flexibility than traditional automation systems. Most notably, AI is seemingly inserting itself into the creative process, where stages in innovation and new product development also leverage AI assistance in product design by generating new ideas, simulating product performance, and optimizing for manufacturability.
While these are the main categories of AI contribution to manufacturing, they are yet a few examples of how AI is transforming the manufacturing industry. As AI technology continues to advance, we can expect to see even more innovative applications in the years to come. This article concludes with some views on what this future may look like.
Company-Specific AI Applications in Manufacturing
The aforementioned examples are general in nature, however some specific examples in industry can also be seen.
- Predictive Maintenance – GE and Honeywell: GE uses AI to analyze data from industrial equipment to predict failures and use AI to optimize maintenance while reducing downtime.
- Quality Control – Toyota and Tesla: Automotive quality oversight is applied all along the production, however final system testing is highly complex and prone to human error. Toyota uses AI-powered vision systems to inspect cars for defects, and at Tesla’s Gigafactory, AI-powered robots are used to inspect, among other areas, batteries for defects.
- Supply Chain Optimization – Walmart and Amazon: Walmart uses AI to optimize its supply chain by predicting demand, managing inventory levels, and improving transportation routes, while Amazon reliance on AI has rapidly grown around warehouse management.
- Robotics and Automation – ABB and FANUC: Among leading developers of AI robotics, ABB’s YuMi collaborative robot uses AI to learn and adapt to different tasks, working safely alongside human workers. FANUC’s robots are focused on precision and flexibility.
- Product Design and Development – Siemens and General Motors: Siemens uses AI in product designs for manufacturability, enabling more sophisticated up-front consideration over manufacturing costs and optimizing performance prior to implementation. GM also uses AI to generate new design concepts and simulate product performance.
As AI technology continues to evolve, we can expect to see even more innovative and impactful applications in the years to come.
AI in Manufacturing: Overcoming the Challenges
While there has been significant success among some companies in applying AI systems, implementation is fraught with challenges. Many companies have struggled to achieve the desired results from their AI projects, leading to skepticism and a lack of enthusiasm for future initiatives. In this blog post, we will explore some of the common challenges that manufacturing companies face when applying AI and discuss strategies for overcoming them.
One of the most critical factors in successful AI implementation is the quality and quantity of data. AI algorithms require large amounts of high-quality data to learn and make accurate predictions. Manufacturing companies often struggle to collect and organize relevant data, and the data they do have may be noisy, incomplete, or biased. In this way, small-scale manufacturers may se limited benefit in the adoption of AI systems, with insufficient benefits to overcome the costs of deployment.
Even if there is sufficient availability of data to make use of AI systems, finding the right people to for implementation is the next challenge. While AI system engineering is increasing in focus and specialization, the on-the-ground team of experts must draw from a wide range of skills in data science, machine learning, and systems engineering. Many manufacturing companies lack the necessary expertise in-house, and hiring external talent is expensive and time-consuming.
A major driver of this challenge is the inherent complexity of integrating AI systems with existing manufacturing infrastructure. Effective AI implementation requires significant changes to existing processes and systems, and when it comes to processes, sometimes the most challenging aspects to change is the human factor. People are often by nature resistant to new technologies, especially if they perceive AI as a threat to their jobs or job security.
From a strategic standpoint, many companies do not have a clear understanding of the business goals and objectives, and there are legitimate concerns among senior management or board-level stakeholders which hold concern that AI projects may not deliver the desired value. Unfortunately, there are many examples of manufacturing companies that have struggled to implement AI successfully. One notable example is the case of GE Healthcare, which invested heavily in AI but faced challenges with data quality, integration, and employee resistance.
To overcome these challenges and achieve successful AI implementation, manufacturing companies must undertake a thorough analysis of AI systems and their cost-benefit case. The foundation of an effective decision rests in the following areas.
- Invest in Data Quality: Prioritize data quality and invest in tools and processes for collecting, cleaning, and organizing data.
- Build or Acquire Talent: Develop in-house AI expertise or partner with external consultants.
- Phased Implementation: Start with small, focused AI projects to build capabilities before tackling larger initiatives.
- Change Management: Develop a comprehensive change management plan to facilitate adoption.
- Clear Business Goals: Define clear business goals and metrics for success.
- Ethical Considerations: Consider ethical implications and implement safeguards to ensure responsible AI usage.
Strategy Support – Filling in the Gaps
Business and strategic consulting, or internal strategic analysis, can go a long way in filling the gaps of knowledge and determining feasibility, prior to engagement of technical consultants.
1. Assess Suitability for AI Solutions
- Conduct an assessment: identify strengths, weaknesses, opportunities, and threats (SWOT) related to AI adoption.
- Identify potential use cases: pinpoint specific areas where AI can add value and align with the company’s strategic objectives.
2. Develop a Clear Business Case
- Quantify benefits: quantify potential benefits of AI implementation, such as increased efficiency, cost savings, or revenue growth.
- Develop a compelling narrative: craft a persuasive business case that highlights the value proposition of AI to help secure buy-in from key stakeholders.
3. Identify Potential Use Cases
- Leverage industry expertise: benchmark use cases and best practices.
- Tailor recommendations: customize recommendations in alignment with strategic objectives and to leverage business core capabilities.
4. Build a Strong Data Foundation
- Assess data quality: evaluate the quality and completeness of a company’s data to identify gaps and areas for improvement.
- Recommend data management strategies: identify processes for data collection and cleaning.
5. Develop a Robust AI Strategy
- Define a roadmap: develop a clear roadmap for AI adoption, outlining key milestones, timelines, and resource requirements.
- Identify talent needs: identify skills and expertise required for AI implementation and system maintenance.
6. Pilot and Scale
- Design pilot projects: design and execute pilot projects to test feasibility.
- Monitor and evaluate: including project management, to monitor and evaluate performance of AI initiatives and facilitate course corrections.
While internal strategy teams can provide valuable support in many areas of AI implementation, there are some steps that require specialized skillsets, particularly with respect to planning and execution of technical implementation, which includes costing of services, infrastructure and software licenses. Another key area where external technical support is needed is with respect to data engineering. Strategic consulting support can help with project management and guidance, and can do so most effectively with cooperation of external experts for technical implementation.
The Future of AI in Manufacturing
As AI technology continues to evolve, we can expect to see even more innovative applications in the manufacturing sector. The integration of AI in manufacturing is transforming the industry in profound ways. Some emerging adoptions include:
- Autonomous Manufacturing: AI-powered systems will be able to operate manufacturing facilities with minimal human intervention. There are already many best practice cases of automated manufacturing.
- Digital Twins: Digital twins of manufacturing plants can be used to simulate different scenarios and optimize operations.
- AI-Driven Supply Chains: AI can enable more resilient and responsive supply chains by predicting disruptions and adapting to changing conditions.
AI in Manufacturing: A Strategic Imperative
AI has the potential to revolutionize the manufacturing industry by increasing efficiency, improving quality, and reducing costs. However, implementing AI solutions can be challenging, requiring careful planning, data preparation, and a strong business case. Internal strategy teams can play a crucial role in helping companies navigate these complexities.
For companies that have not actively pursued AI solutions for improved efficiency, quality and performance, now is the time to get on the bandwagon. A 2023 McKinsey study determined that AI technologies can automate tasks that absorb between 60% and 70% of workers’ time today, and a another McKinsey survey found that 63% of respondents in manufacturing had adopted AI in at least one business function, up from 51% in 2019. By conducting thorough assessments, developing clear business cases, and identifying suitable use cases, strategic analysis helps fill the gap for companies exploring their degree of readiness for AI adoption, and enables a strong data foundation.
While small strategy consultancies may not have the in-depth technical expertise to implement AI systems themselves, they can provide strategic guidance, industry knowledge, and practical support. By partnering with consultancies or leveraging internal strategy teams, manufacturing companies will increase their chances of successful AI implementation and unlock the full potential of this transformative technology.