Henry Ford didn’t invent the automobile. It debuted more than 20 years before the first Model T rolled out of his Rouge Plant in Dearborn, Michigan. 

What Ford did invent was the concept of the manufacturing assembly line. And that’s what gave the automaker the upper hand in a nascent market. 

Today, the level of automation at the Rouge Complex is eons ahead of what Henry Ford could have imagined in his wildest dreams. The foundation’s still there, but it’s been greatly modernized. 

A chassis moves from one station to the next, where workers are equipped with the tools and skills to execute precise actions. No single worker performs more than a couple of steps within the assembly line.

One of the last stations at the Ford plant consists of a chamber where cameras take thousands of pictures of the newly assembled vehicle, and quickly bring them to the screen in front of the worker, so that the smallest imperfection can be detected. 

Throughout the facility, humans and robots are working together. What’s not immediately evident, however, is how artificial intelligence is aiding the modern-day manufacturing process — unlike that of any automation in the past. AI is the ideal tool to iterate and achieve the highest levels of speed and accuracy in manufacturing.

Spurring Technology Investment

Advances in manufacturing technology in the last three years have been primarily driven by the passage of three key legislative acts:

The CHIPS and Science Act allocates funds to support research and development in the semiconductor space. It’s crucial for the advancement of artificial intelligence platforms, and reducing the dependency on external semiconductor developments by the U.S. market.

The Infrastructure Investment and Jobs Act incentivizes the development of a stronger and more qualified workforce, promoting the acquisition of new skills to support advances in technologies such as AI.

The Inflation Reduction Act is promoting clean energy initiatives and a more sustainable approach to increased productivity in the sector. AI makes it easier for businesses to proactively detect defects and reduce rework, which in turn reduces potential for machinery downtime and energy consumption due to inefficiencies in productivity.

Conventional manufacturing processes often result in material waste due to inefficiencies in the process. This is largely driven by manual steps that lack precision, and poor planning methods that can lead to a surplus in product or loss of sales because of lower production levels versus market demand.

Additionally, the lack of standardized processes and dependence on manual labor makes it harder to scale operations in an efficient way. Variability in product quality and production times can lead to significant reworking and inconsistent outputs. These problems don’t stop in the production line; they extend to areas such as procurement and inventory management, among other key functions.

A simple exercise to identify challenges you could be facing, along with opportunities for productivity gains, is to ask the following questions:

• Where are the biggest sources of waste in our current production process? Do we have excess inventory? What is the number of defects we have identified per million products? 
• How could we improve our efficiency by redesigning plant layout? Do we have opportunities to streamline transportation and reduce waste by reducing movement? What steps in the process could be benefitted by automation?
• Do we have specific key performance indicators and metrics that give us visibility of the bottlenecks and the low-hanging fruit to deliver higher efficiency and reduce waste? What has been the total downtime of the plant over the last two years? How much of it has been driven due to repairs that could have been avoided by proactive maintenance?

Successful Applications of AI

AI has been a clear driver for improvements in productivity in several key areas:

Predictive maintenance and reduction of downtime. Siemens implemented an AI-based predictive maintenance system in its manufacturing processes. It embedded sensors that relay telemetry data about the machinery in the production process. The resulting data is analyzed using AI algorithms to better predict when a machine is likely to fail. This allows Siemens to perform maintenance before a failure occurs, and significantly reduce unplanned downtime and improving overall equipment efficiency.

Quality control and reduced rework. Century Plyboards, one of the leading producers of plywood, laminates, doors and veneers in India, had significant issues in the quality of its products, such as uneven core sheet density, because manual checkpoints were finding it hard to locate and address the specific cause. The solution involved capturing real-time images of the core sheets as they’re fed into the production machines, using AI models to determine with high precision if the sheets in question have flaws.

Workplace safety. The European Agency for Safety and Health at Work identified a number of case studies of companies that implemented AI for the benefit of worker’s safety and wellbeing. In one case, a company had significant challenges due to the cognitive load on workers and the repetitive nature of the tasks they were performing when manually inspecting parts. By implementing an AI-based system for inspection of final products, the company was able to automate many repetitive tasks. The result was a safer and more efficient work environment.

Following are some questions that can highlight areas of opportunity in similar scenarios:

• Where do we have bottlenecks or inefficiencies in our operations? 
• Are we utilizing our resources in the most effective way possible?
• Where can we improve our position in our supply chain? Do we have models on demand forecasting and how can we use it to prevent being impacted due to disruptions in our supply?

GE made news in 2015 with the release of its own industrial internet of things platform called Predix. It was an attempt to provide something similar to what Siemens did on its own: create a framework that would collect information from sensors in machines in production facilities, provide analytics and predictive models to drive operational efficiency, and optimize maintenance costs. 

Unfortunately, GE’s approach faced major challenges, including the complexity of integrating many different systems at scale, and significant change-management challenges (a typical failure point of most digital transformation initiatives). 

GE invested over $7 billion in its development, but the lack of clear objectives, limitations in platform scalability and poor development of internal capabilities to support the platform proved to be hard barriers to its success.

Some questions to consider to identify key challenges in an organization would include:

• Do we have clear visibility of the end goal? What is the compelling event that’s driving our organization to look to AI as a solution?
• Do we have the necessary skill sets in our resources to support a digital transformation initiative of this magnitude?
• Do we have visibility of the data and systems that will be required, and any potential technical, in terms of integration or long-term support?

AI is transforming the manufacturing sector by significantly enhancing productivity, quality and safety. Examples like Ford and Siemens demonstrate the way AI-driven systems can revolutionize manufacturing by enhancing processes and controls on the production line — and doing it at scale and with greater precision.

The landscape is primed for organizations to invest in and develop these capabilities. To ignore the AI revolution is to risk being left behind. Embracing AI is crucial for staying competitive in the evolving industrial landscape.

Jose Paez is director of solution strategy for Pricefx.