Industrial digital twins are gaining momentum in the manufacturing and infrastructure industry. Though industries have yet to reach a consensus on the definition of a digital twin, in our previous coverage, we have showcased use cases involving "design twins" and "operations twins." The manufacturing industry uses design twins for planning operations, virtual commissioning, AR/VR applications for training, and management of infrastructure projects. Similarly, operations twins enable monitoring of a machine's health during operations and predictive maintenance. However, the use of digital twins focuses on the specific stages of the product life cycle. For instance, design twins cover early phases of the product design, and operations twins bring much-needed process visibility during the production phase.
Digital maturity and the drive for sustainable products call for digital twins operating throughout the product life cycle. A digital twin for closed-loop engineering combines data from the physical asset in real-time for potential design upgrades in the future. Though most use cases focus on the design and production phases, enabling digital twins throughout the product life cycle offers distinct advantages, such as:
- Improved product design by capturing real-time feedback for design improvements and faster new product development.
- Improved process visibility and easier management of maintenance and service, especially in industries where traceability is key during the process, such as aerospace or pharma.
- Better management of the end-of-life phase with an integrated digital twin given the growing need for sustainable product development and push for the circular economy.
Moving ahead, we expect to see a rise of digital twins for the closed-loop engineering use cases. There are three key points innovators should consider while working on digital twins for closed-loop engineering applications:
Integration of physics-based models & data-driven models:
Digital twins must be connected to the physical asset in real-time and combine physics-based models and data-driven models. Physics-based models are predominantly used in the design phase for high-fidelity simulations, whereas data-driven models are employed in the operations twin. Closed-loop engineering requires the integration of both. However, it requires a platform that supports conventional computer-aided engineering (CAE) tools and sensor data connectivity, analysis, and the ability to implement machine learning and AI algorithms. Recognizing the need for integration, large players are looking for companies that complement their existing digital twin portfolios. For instance, Bentley Systems acquired Seequent, a geological modeling company, for $1 billion; on the other hand, Hexagon AB acquired CADLM to expand the reach of design twins. There are very few startups offering digital twins as a product that can integrate design and operations twins. Startups have a limited role to play and target specific industries, for example, Akselos in the O&G industry, Seebo in the chemical industry, and Mevea for hydraulics applications. The large players have an advantage with reconfiguring existing design twin platforms to accommodate operations twins. Hence, platform providers Siemens, Altair, Ansys, PTC, Hexagon AB, and Bentley Systems are well-suited for the closed-loop engineering integration. However, companies should note that accepted standards for data models and interoperability among platforms remain a distant goal.
Cloud vs. Edge vs. on-premises deployment:
Large-scale projects involving multiple organizations and geographies must adopt cloud-based solutions. Cloud platforms like Altair One offer high-performance computing solutions for high-fidelity simulations. Though cloud deployment presents an excellent case for the overall management of digital twins, time-critical applications require digital twins at the edge. Similarly, many maintenance and repair AR applications access a part of the digital twin on AR hardware. However, edge deployment is still in the early phases of development, and some open-source projects plan to explore the feasibility of digital twin management at the edge. Organizations working on sensitive data and others worried about losing competitive advantage due to data sharing may opt for on-premises deployment. On-premises deployment may be more secure and beneficial in a few use cases, but we see a trend toward a hybrid model combining the edge and the cloud for digital twin deployment.
Management and evolving business models:
Closed-loop engineering projects are expensive and may involve multiple teams from various organizations, which raises an important question about management and value extraction. As the digital twins are passed along the value chain throughout the product life cycle, the partner extracting the value may not be the same one that initially invested in the digital twin. Moreover, as ownership is transferred throughout value chains, so will the inherent liabilities. There is a possible role for a third party that can manage digital twins throughout the life cycle. Additionally, increased visibility and continuous monitoring with closed-loop engineering digital twin deployment will enable new business models like machine-as-a-service or product-as-a-service.
Although it may be tempting to embark on a digital twin project, lacking standardization of information models and limited interoperability remain significant barriers to broader adoption of digital twins for closed-loop engineering applications. Recognizing the need for accelerated standard development, the Industrial Internet Consortium, Plattform Industrie 4.0, Digital Twin Consortium, and Industrial Digital Twin Association are working on open standards for digital twins. In addition to establishing standards, it is essential to evaluate the digital maturity of the partners involved in digital twin development for closed-loop engineering. Going forward, as digital twin platforms mature, along with general digital maturity, challenges regarding ownership and management of digital twins need to be addressed. Digital twins for closed-loop engineering applications are most suitable for high-value complex assets that undergo frequent operational changes or need upgrades throughout product life cycles, such as aircraft engines, so expect to see initial deployment in those areas.