By Anindya Chatterjee, GVP, domain value engineering and industrial AI, ABB
Industrial operators currently face pressure on several fronts—economic uncertainty, energy-related challenges, and a “talent crunch” caused by an aging workforce and a rapidly evolving job market.
These challenges are particularly affecting asset-intensive producers with complex processes, given the scale and nature of their operations. For example, many operators suffer from a lack of systems and processes and are unable to determine where losses are happening. An inability to assess the health of an asset means that engineers cannot avoid incipient failure.
However, the power of digital offers some welcome cause for optimism, given the possibilities these technologies enable for tangible, real-world operational improvements, not least in terms of things like predictive maintenance; predictive analysis of equipment that enables engineers to predict incipient failure of assets can significantly improve overall equipment effectiveness (OEE).
Moreover, introducing sophisticated predictive maintenance means we are able to learn from the past, compare it with the present, and get better at identifying faults before they arise. Many symptoms from machines go undetected because of operator limitations—more often than not, this is an issue of time and resources. Digital technologies offer real-world efficiencies through alerts, data and early-warning about faults…quantitative assessment rather than a solely human qualitative assessment.
Many operations face difficulties aggregating and analysing the vast amounts of data produced, as well as under-utilized assets. They walk a fine line between ease of use for digital solutions, while maintaining secure operations.
However, it is possible to use what you have, better, and scale this across an entire operation—building on existing digital technology and leveraging automation that is already controlling industrial processes, just doing so in a smarter, faster, more secure way.
This can have a major impact on efforts to predict issues and prescribe actions that can help operators better utilize their assets and fine-tune production processes—overall building more resilient, more productive operations, whether that is in hydrocarbon processing, manufacturing or pulp and paper.
It all starts with finding the particular, industry-specific pain point within the business and finding solutions to collate and contextualise data across each data source. These data sources include OT, IT, and engineering technology (ET), in order to gain greater insight about operations—and to identify actionable steps that can be taken to make improvements.
For example, an ABB customer in the process industries faced high operational losses of around 5% and 15 days of asset downtime, which had significant impact upon the paper mill’s revenue. We deployed our industry domain knowledge to quickly account for where and why these losses were happening—automated loss identification—and began more sophisticated tracking based on root-loss analysis. We then introduced predictive-maintenance tools and analytics to better predict nascent failures in these assets.
Domain knowledge is the key to successful and swift implementation, with this knowledge coded into the solution. It is crucial for operators to avoid the temptation of generic IIoT analytics and AI platforms that do not have industry-specific knowledge embedded beyond core, critical capabilities.
With built-in domain-specific coding, time to implement is vastly reduced and operations can start to see the benefits far sooner, as months of tailoring and programming (post-install) is removed from the equation.
Experienced engineers and data scientists who have a deep knowledge of their industries—and the operations they are responsible for—are crucial to refining, adapting and learning what works best. They can steer the industrial AI and analytics accordingly. Combining the two forces can make a world of difference to operations.
Pre-code domain expertise in software + data =operational excellence.
Insights + experience = smarter decisions.
Optimization ultimately means more productive, efficient and sustainable operations with digital software that is constantly learning and, thus, helping you meet and adapt KPIs as you go.
Given the strong productivity gains that can be realized through the integration of AI, there is ample opportunity to gain the benefits of industrial AI and analytics to build a competitive advantage, despite the many challenges currently facing the industry.
Consider there five tips:
- Work with a digital partner who knows your sector
- Evaluate IIoT digital solutions that have been built for your industry or with in-build domain expertise
- Ensure that the fundamentals are in place within your systems (e.g. cybersecurity patching)
- Maintain an ongoing training schedule for teams
- Constantly iterate, adapt and grow with the AI / machine learning capability
