Deploying AI for manufacturing success. By Alexander Rinke
The age of automation is upon us, with Artificial Intelligence (AI) transforming every aspect of our lives. At home, asking Alexa to order goods and give us the weather forecast is quickly becoming the norm, and at work, robot colleagues, or co-bots, are quickly becoming a common fixture in many factories and production lines. Similarly, the UK government is starting to wake up to AI’s potential to deliver a ‘major boost’ to the economy. In 2017, it announced an initiative to invest £1 billion into the sector as part of its Industrial Strategy. Manufacturers such as Siemens and Chart Industries are already capitalising on AI to make operational improvements and drive efficiencies in their businesses. But while the benefits of automation are clear, and many manufacturing businesses realise they must now start thinking about AI initiatives, the real challenge for many lies in understanding how and where it can actually be deployed effectively.
Tackling AI head-on
Looking at internal processes and the way their organisations operate is one way for manufacturers to make improvements. However, many remain in the dark about where the roadblocks are in their operations, and how to find them. They often don’t know where to start in terms of which of their business processes are ready for automation, if their processes are too complex for automation, and how to monitor bots after they have been deployed. This is a common theme amongst many manufacturing businesses, and more guidance is needed if they’re to use AI to see legitimate results and remain competitive.
Another challenge that manufacturers face is changing the perception of automation within their organisations. A narrative has emerged around the disruptive nature of AI, with many fearing that jobs are at risk of being replaced by bots that are more skilled, and don’t tire as easily as humans. In reality, AI will create as many jobs as it will replace, according to a new PwC report. The manufacturing industry will naturally see more displacement than other industries, with lots of time-consuming, repetitive, and menial tasks already being carried out by machines. Doing so frees up human brains to focus on more complex tasks, enabling workers to concentrate on strategy, creative innovation and problem solving; which in turn, will help create new roles that don’t exist yet.
Knowing where to start
For those manufacturing businesses that are ready to deploy AI, it’s important to remember that every successful automation initiative starts with a solid analytics foundation. The ‘magic’ of AI is built on pattern recognition and machine learning, which requires data to learn from. Common concerns for manufacturers are that they don’t have data that’s ready for AI, they aren’t collecting the right data, and/or they’re collecting it from too many separate sources. But technology has evolved to a point where algorithms are powerful enough to help enterprises use the very data they already have stored within their own systems.
One such technology is process mining, a new category of big data analytics that helps manufacturing businesses get a clear view of all the activity that’s currently happening within their organisation. It uses the digital traces left behind by every IT-driven operation to reconstruct the as-is scenario and provide complete transparency into how business processes are operating in real life.
Using this insight, manufacturers can easily identify which of their processes are ready for automation, and whether these adjustments are justifiable with ROI. Getting a complete picture of business processes is critical to helping manufacturing companies develop a more targeted approach to automation, and process mining can help here. Combined with machine learning, the technology becomes a powerful tool for businesses, by providing prescriptive recommendations for process improvement, and alerting users to previously undiscovered opportunities for greater efficiency.
It is no surprise that businesses and world leaders alike are starting to sit up and take notice of AI and the potential impact it could have. The benefits are set to be enormous, with the promise of new jobs and a significant boost to the economy, in turn, enabling businesses to remain competitive on the global stage.
Of course, the power of AI stems from the capability to understand a business’ challenges more completely and to make changes as a result. It is unwise to dive headfirst into an automation initiative without having an objective, so any manufacturer considering rolling out AI must first identify its intended purpose. If the business is unprepared or unwilling to make changes based on the potential that AI can unlock, then the whole initiative is wasted.
Alexander Rinke is co-founder and CEO, Celonis. Celonis is the New York- and Munich-based leader in business transformation software, turning process insights into action with the process mining technology it pioneered. Its Intelligent Business Cloud allows organisations to rapidly understand and improve the operational backbone of their business. Companies around the world including Siemens, GM, 3M, Airbus and Vodafone rely on Celonis technology to guide action and drive change to business processes, resulting in millions of dollars saved and an improved experience for their customers.