AI-driven transformation improves operational efficiency—but it requires leaders to strengthen their foundations to take full advantage
In 2025, AI is rapidly expanding and changing industries like manufacturing. Manufacturers expanded last year for the first time since early 2023, with many banking on the potential of AI automation to power growth. While that expansion has cooled in the face of a potential economic slowdown, the industry still has many opportunities this year—as well as hurdles to navigate.
AI offers many benefits in manufacturing. In fact, AI is already changing many aspects of the field, including product development, quality assurance, equipment maintenance, and supply chain management.
We’ve collected data and trends, and synthesized our own industry-insider perspectives below. We’ve outlined the major challenges facing manufacturers, the opportunities for AI, and steps for enterprises to get started with targeted, high-ROI use cases.
Key Challenges in Manufacturing
Manufacturing—like every other industry—faces difficulties this year. Amid economic uncertainty, shifting consumer values, and the fast-paced advancement of technology, forward-looking leaders are strategizing for the following challenges.
- Supply chain disruptions. In 2024, there was a 38% increase in global supply chain disruptions compared to the previous year. This marks a major departure from just a few years ago, when disruptions were around 5%. Geopolitical tensions, trade conflicts, and natural disasters contributed to raw material shortages, increased costs, and delayed production schedules. With these memories fresh in manufacturers’ minds and new proposed tariffs introducing additional stress, many are bracing for continued disruption and seeking ways to build resilience.
- Skills gaps and labor shortages. Manufacturing is facing personnel issues, especially in skilled positions such as machinists, engineers, and technicians. This creates major issues that will continue throughout 2025 and into the years to come. Projections for 2025 peg a potential shortfall of 1 million workers. There’s a major mismatch between skills required by manufacturers and those possessed by the available workforce, especially as new technologies demand expertise in AI, robotics, and data analytics. By 2030, a lack of skilled labor will lead to an estimated 1 million unfilled jobs in manufacturing.
- Sustainability. Currently, the manufacturing sector accounts for one-fifth of global carbon emissions and 54% of the world’s energy usage. Manufacturers are under growing pressure to reduce their environmental impact, comply with regulations, and meet consumer demands for sustainable products. This creates pressure to invest in green technologies, improve energy efficiency, and redesign products and processes for sustainability. Balancing these initiatives with cost controls and competitive pricing is a significant challenge.
- Managing digital transformation. AI and associated technologies are poised to transform manufacturing operations, offering the potential for the industry to establish smart factories, AI-powered supply chains, and even predictive maintenance processes. Yet, the skills gap in manufacturing, coupled with a prevalence of legacy data management systems and insufficient change management strategies, creates growing pains on the road to transformation.
How AI Is Already Solving Problems in Manufacturing
AI is impacting the manufacturing sector and will cause more drastic changes over the next five years. Some AI-driven transformations that are already taking the industry by storm cover critical areas like resource and supply chain management, production, and maintenance.
Optimized resource allocation
AI is driving efficiencies that reduce costs and increase profit margins. AI can save businesses money by optimizing staff schedules and monitoring energy use, finding new ways to reduce consumption. Simultaneously, AI can support predictive maintenance, allocating resources appropriately during high-demand times, and flagging opportunities for maintenance when demand is lower. This trims operational costs and allows manufacturers to make the most of high-volume periods.
Smarter customer service
AI is changing how businesses interact with their customer bases. About 80% of business leaders use intelligent AI-powered customer service bots to reduce the number of customer interactions fielded by their human workers. In this context, AI minimizes the need for human personnel to field basic queries and helps customers get the information they need quickly.
AI-powered supply chains
Manufacturers can leverage AI to analyze shipping documents using natural language processing (NLP), providing managers with valuable information to improve processes within the supply chain. AI can also be used to track inventory, reduce errors and waste, and power predictive analytics to make supply chains more environmentally sustainable. Manufacturers can leverage AI and machine learning (ML) to optimize truckloads, predict the most fuel-efficient delivery routes, and reduce product waste in the marketplace.
Robot-powered production
We can’t talk about new technology and AI without discussing robots. Robots have been performing manufacturing work, such as assembling products and spot welding, since the 1960s. Some estimate that 20 million manufacturing jobs can be either downgraded or eliminated due to automation. That might sound scary—but it represents a potential economic boost for manufacturers that will allow them to invest in upskilling their workforces to meet the skills gap.
The Human Side of AI in Manufacturing
AI isn’t only changing the actual manufacturing processes and products being made. It’s also changing workers’ roles. No longer are they doing tedious tasks; software is now doing most of the menial, repetitive work.
Research shows that AI automation is slashing how much time human workers are spending on repetitive tasks by 45%. It can save between 10 to 20 hours per week, which has the potential to improve productivity and engagement.
If workers aren’t doing repetitive tasks, what are they doing instead? According to the National Association of Manufacturers (NAM), workers will learn new skills to focus on tasks involving problem-solving, complex reasoning, and oversight of AI-supported processes and machines. Companies should start updating job descriptions to reflect the skills they need from employees in the coming years.
Here are our notes on the key skills manufacturers should seek in their workforces:
Advanced technical manufacturing skills:
- Operating, maintaining, and programming advanced manufacturing equipment, including CNC machines, 3D printers, and robots
Data analysis and management experience:
- Analyzing Industrial Internet of Things (IIoT) data
- Understanding of data management practices and principles
Software development and engineering ability:
- Designing and maintaining manufacturing software systems
- Knowledge of and ability to work with digital manufacturing processes
Cybersecurity expertise:
- Protecting sensitive data and manufacturing items from cyber threats
Knowledge of systems engineering and integration:
- Designing, integrating, and managing complex systems, including hardware and software components
Advanced supply chain management:
- Understanding of complex global supply chains and logistics
- Integrating new solutions into the tech stack to facilitate supply chain optimization
Soft skills:
- Problem-solving, teamwork, creativity, adaptability, and the desire to learn and improve continuously
The manufacturing sector will also likely need to create and hire new positions related to AI, including data scientists and ML programmers.
Challenges of AI adoption in manufacturing
In addition to general manufacturing challenges, the industry also faces specific challenges associated with adopting AI. These include:
- Data quality and availability: Data in manufacturing is often unstructured and inconsistent, making it difficult to use in training and deploying AI models.
- Silos and disconnected systems: Data stored across disparate and disconnected systems makes it difficult to consolidate and share with engineers designing AI models.
- Integration with legacy systems: It is common in manufacturing—and in other industries—to rely on legacy systems (both hardware and software) that may not be compatible with AI solutions. This means that introducing AI can require costly upgrades.
- Data security: Manufacturers handle lots of sensitive data. Implementing AI to process sensitive information about manufacturing processes and personnel requires a robust cybersecurity architecture and culture of data security—which many manufacturers using legacy systems don’t have.
How to Adopt AI in Manufacturing
If your manufacturing company is just starting to incorporate AI into company operations, here are some foundational tips to help you get started.
Focus on data and purpose-built, vertical solutions
Good AI models require good data and people with the right expertise to work with them. But getting data into usable condition and hiring talent to build AI models is expensive. An alternative to a custom-built AI solution is a data-centric vertical AI platform. For example, an automated anomaly detection tool could replace or augment human workers tasked with quality control in factories. Platforms like Tulip also empower manufacturing engineers to build and customize the tools they need in a low-code interface for analysis and response to operator feedback.
Prepare for organizational redesign
The manufacturing industry will have a widespread need for new roles and operating models. If companies rely on AI-generated insights, they need a human layer that systematically governs data quality and automation results. As AI changes how people work, we must adapt organizational structures to match. Some new organizational structures with AI include hubs, spokes, and execution teams. Hubs are central executive functions that align strategy with analytics and execution. Spokes are the strategic business units, managed by human teams, where the AI functionally operates. Execution teams collect data, train models, and launch AI-powered capabilities at the spokes level. This organizational redesign focuses labor on strategy and work that only humans can do, while AI takes over more transactional, repetitive tasks.
Start with a few use cases
Find a handful of use cases that are both valuable and feasible, meaning they can be prototyped and released fairly quickly, in a matter of weeks or months—not years. Consider applying generative AI and other technologies to targeted initiatives to learn, develop skills, and secure early wins that can build organizational momentum and gain buy-in. To get quick wins, work with a small team to design and deploy the minimum viable solution to validate AI’s potential in business operations.
Quantify ROI
Measuring ROI in AI initiatives ensures that organizations are strategically leveraging these technologies to generate measurable value and benefit their bottom line. This strategic approach ensures that AI adoption is a value-generating investment. There are a number of solutions that measure the ROI of digital transformations in user-friendly ways. These are worthwhile investments for manufacturers that want to truly use AI practically and sustainably.
The Lifecycle Management Approach to AI Adoption
Since there are many use cases for applying AI within manufacturing, it’s important to take a deliberate approach to AI projects so you can deliver tangible business benefits quickly. The approach we recommend is the AI lifecycle. This approach starts with defining a few concrete problems first before identifying potential AI solutions. Then, a small, qualified group will evaluate all proposed AI ideas and narrow the list to a few for prototyping. Decisions are made based on the use cases’ value, feasibility, and risks.
This approach emphasizes the importance of developing prototype or pilot solutions first to validate the value hypotheses of AI use cases (“Will this AI solution increase this business KPI by X percent?”). The feedback from the pilot is used to decide whether an AI solution is ready for wider implementation, needs refinement, or should be discontinued altogether.
Once decision-makers identify which prototype solutions are worth implementing more widely, a small, interdisciplinary team thinks through all potential implementation hurdles and addresses them to maximize the chances of the solution gaining traction. Once a new AI solution is rolled out, a team monitors its usage and business benefits and makes adjustments to ensure it delivers the expected business value.
This approach helps companies sustainably implement AI solutions at scale. It also allows you to adequately manage risks and ensure that AI projects are executed efficiently.
Win Big with AI
Ready to drive business value with AI? Try Shibumi’s AI Value Accelerator. This software solution helps companies govern AI projects, manage them through a lifecycle, and drive measurable business value. Additionally, our centralized AI Idea Hub helps leaders source ideas centrally from a secure platform, while our prioritization engine helps decision-makers leverage out-of-the-box reports to prioritize the AI opportunities that will provide the most ROI.
Need a more tailored approach to get started? Reach out to one of our staff experts for a free consultation on how to implement and make the most of AI in your manufacturing business.