AI is driving digital transformation in engineering

Artificial intelligence (AI) is driving digitalization when it comes to transforming the engineering sector.

By Sonali Singh and Sandeep Mohan November 4, 2022
Courtesy: Brett Sayles

Digital transformation insights

  • Engineering firms are using digital transformation to consolidate and connect portfolios of engineering software and technology to help companies make better and more informed decisions.
  • AI-driven digital twin technology captures real-time data of the asset once it is in operation and can bring a lot of value to a company.

As digitalization ramps up across all industrial sectors, so does the need to rethink how products, services and engagement strategies are implemented. Digitalization has become even more critical today as companies grapple with the dual challenge of meeting both profitability and sustainability targets. AI-driven digitalization can help address this dual challenge.

For many years, the different engineering disciplines operated alone, focusing only on their specific fields, and that worked. However, market dynamics have now forced process engineers to look at the bigger picture and to optimize their processes by integrating different engineering disciplines and to improve collaboration.

Enhanced workflows that enable better communication and sharing of information between process engineers, project engineers, estimators and safety and energy specialists deliver the potential to capture the opportunities offered by collaboration, which can help uncover opportunities to improve both profitability and sustainability performance of the plant.

In terms of digital transformation itself, most initiatives are being undertaken by engineering firms are to consolidate and connect portfolios of engineering software and technology in support of new, streamlined, digitalized workflows spanning departments, disciplines and offices. Anticipated business benefits include shorter cycle times, lower cost, and higher-quality designs, as well as fewer errors and problems encountered during design, construction and handover.

Meanwhile, operating companies have always struggled closing the gaps between the operational behavior predicted by theoretical engineering models and actual real-world operations. Traditionally, this has been addressed by manipulating ‘tuning factors’ for engineering models, which requires many years of experience. It also demands continual attention as the tuning factors have to be changed from time to time. With more new generation engineers unwilling to spend years in one particular job function, operating companies are finding it hard to develop and retain the required expertise and has prevented many operating companies from taking advantage of the advanced solutions available.

AI opportunities for engineering

Today, a growing number of engineering firms are extending their digital initiatives into new areas to take advantage of escalating business opportunities. Artificial intelligence (AI) often lies at the heart of these opportunities.

The use of AI-driven digital twin technology, in particular, can bring high value to the engineering company’s clients, as it captures real-time data of the asset once it is in operation. Additionally, being in a position to leverage digital design and engineering data during handover provides greater potential for offering value-added services during operations and maintenance, making organisations less reliant on capital projects alone.

AI can also play an important role in supporting and outlining an engineering approach to a project, the constructability of a design, and the planning of how materials, equipment and labour are organized. This early planning approach has been proven to reduce costs and speed up schedules.

AI has had an even greater impact on operating companies. AI technologies can now easily read into the plethora of operations data already available with most operating plants and determine equations that automatically computes the tuning factors that match the engineering model predictions with real world operations. This saves the operating companies from their dependence on third-party consultants to maintain the accuracy of engineering models and it also drastically reduces the frequency of updating these models.

This has had profound impact in both sustainability and profitability of a plant. For example, it enabled a U.S. refinery to get unprecedented accuracy in simulating and predicting the performance of its FCC reactor. FCC is one of the largest CO2 emitters and one of the most critical units in producing high value products in the refinery. These highly accurate low maintenance AI supported models helped the refinery uncover insights to improve operations as well as in improve the accuracy of planning operations. This enabled the refinery to improve the production of high value products while ensuring it met emission targets.

The use of first-principle models for defining and predicting a project’s performance and its outcomes is standard in process industries. However, there are some processes that are more difficult to predict. Often these are managed through less-precise techniques such as operator experience or rules of thumb, but this can lead to less than expected performance levels.

However, AI can simulate thousands of design options which very quickly narrow down the options that not only best meet the owner requirements, but are the safest, most environmentally-friendly and most cost-effective. Multi-case analysis is a capability that offers engineering firms the opportunity to transform the way these early decisions are made.

Previously, engineers would define these critical parameters using limited data from an equally limited number of operating cases and conditions.

Multi-case analysis helps to optimize early design decisions, based on the consideration of hundreds or even thousands of operating conditions and cases. Leveraging AI and high-performance computing – either in the cloud or on a desktop – allows designers to rely on a significantly broader set of data to adjust and fine-tune their designs.

From the many different grades of crude oil to varying ambient weather conditions, this improvement in understanding how a potential design would perform in real conditions can result in improvements across the board – from construction materials; the size of equipment; to the type of utilities and even the location of the plant. These decisions will often have a major impact on the plant’s capital and operating costs, its risk analysis, as well as the overall fit for its intended purpose. A major global technology provider for natural gas liquids (NGL) processing plants, reported this technology helped it to halve the design time required to optimize the plant design across numerous operating conditions along the lifetime of the plant.

Multi-case analysis is undoubtedly a key area of focus for engineering firms using AI-driven digitalization today. However, in speaking with customers across all regions, the highest priority area to be addressed under digitalization is the consolidation of engineering software and technology portfolios, followed by digitalization of remaining applications and business processes. The ability to find and re-use data across the organization and eventually across the ecosystem of vendors, sub-contractors and consultants is critical.

There is so much to be gained (some companies estimate there is an opportunity for double-digit improvement in engineering and estimating productivity alone) that those who do not move forward risk being less competitive in the future.

One major drawback of pure AI-based solutions is the results depend on the quality of the data it has access to. In the case of operating plants, instrumentation can become faulty leading to “bad” operations data stored in the plant data historian. A purely AI-driven solution will not be able to reliably detect these discrepancies. A hybrid solution that combines the power and agility of AI with the checks and balances of a first principle based rigorous engineering technology can effectively address this challenge and is ideal for the operating companies.

Positive prospects for engineering, digital transformation

Looking ahead to the future of the engineering sector, digital transformation will inevitably accelerate and bring about significant improvements. While the path of change was already firmly established prior to the arrival of COVID-19, the pandemic has acted as a catalyst to speed this up. For engineering firms to succeed it is imperative to adopt the advanced technology that is now available – including AI tools.

This will ensure organizations achieve operational efficiencies across the end-to-end value chain, giving them a competitive edge. That, in turn, will position them well as they look to migrate beyond Industry 4.0 to the rapidly-emerging fifth industrial revolution, which takes the existing paradigm one step further by highlighting operational excellence and innovation as key drivers for change.

– This originally appeared on Control Engineering Europe’s website. Edited by Chris Vavra, web content manager, Control Engineering, CFE Media and Technology,

Author Bio: Sonali Singh is vice president, product management at AspenTech; Sandeep Mohan is senior product marketing manager at AspenTech.