Digital twins have come a long way since their inception, with advances in technologies such as CAD, IoT, cloud computing, and big data analytics making them easier to adopt and much less expensive. The initial release was a simple simulation-based event, where organizations created virtual versions of physical structures to test them under various load situations. Its main advantage is that the model can identify potential risks without actually investing in physical prototypes; Virtual models provide unlimited information about the product and its design in a variety of operating scenarios.

According to the composition analysis of digital twin ecology by China Electronics Standardization Institute, the digital twin ecosystem can be mainly divided into basic support layer, data interaction layer, model construction and simulation analysis layer, common application layer and industry application layer.

Artificial intelligence technology is mainly used in the simulation analysis level, in the simulation analysis layer, according to the "Digital Twin Application White Paper" released by the China Electronics Technology Standardization Institute, how to achieve value extraction through efficient mining methods in large amounts of data is one of the key problems of digital twins. Digital twin information analysis technology, through AI intelligent computing models and algorithms, combined with advanced visualization technology, to achieve intelligent information analysis and auxiliary decision-making, to realize the monitoring and visualization of physical entity operation indicators, the automatic operation of model algorithms, and the online rehearsal of the future development of physical entities, so as to optimize the operation of physical entities.

As one of the underlying key technologies of the digital twin ecosystem, the integration and evolution of artificial intelligence and digital twin also runs through these four stages.

In the 20s of the 80th century, three-dimensional design software represented by CATI was born, upgrading product design from 2D to 3D, bringing "what you see is what you get" technology. From the size, material, appearance, realistically express the geometric appearance of the product. The virtual twin at this stage is primarily a geometric appearance twin. Based on artificial intelligence algorithms, it can perform cluster analysis on the enterprise parts library, realize automatic classification and retrieval of parts, improve the standardization level of parts, reduce maintenance costs and procurement costs, and improve product quality.

In the 20s of the 90th century, product design entered the era of digital prototypes. Digital mockups carry not only the geometric appearance knowledge of the product, but also the multidisciplinary and multi-expertise inherent in the product. The virtual twin at this stage evolved into a multidisciplinary multidisciplinary twin. Artificial intelligence technology combined with multidisciplinary multidisciplinary twin technology has enabled computer-aided design to evolve into generative design, also known as generative design or cognitively enhanced design. For part design, machine design and processing methods, artificial intelligence technology will combine the selection of the optimal design results and production methods through algorithms.

In the late 20s of the 90th century, the concept of the full life cycle of products was introduced. AI technology has a wide range of applications in all aspects of the product life cycle. In the demand analysis stage, the web crawler technology is used to obtain user voices from various forums, and the market demand is defined with the help of semantic analysis and data insights. In the design phase, the design cycle is shortened by using cognitively enhanced design by clustering the enterprise parts library to improve part reuse. In the manufacturing operation stage, through supply chain optimization, workshop logistics optimization, APS, etc., artificial intelligence technology is integrated to improve efficiency and reduce costs.

In the 21st century, fierce market competition made companies pay more attention to how to bring the best experience to consumers, and the scope of virtual twins expanded from within the enterprise to the context in which the product was used. Taking a car as an example, the car must drive on the road, in order to make the driver and passengers get the best driving experience, in addition to the car itself, driving in the city must also build a virtual driver, road, urban environment.

According to research firm MarketsandMarkets, the global digital twin market is expected to grow at a CAGR of 58% to reach $2026.2020 billion by 31 from $482.<> billion in <> as demand from the healthcare and pharmaceutical industries increases. While specific data for India is not yet available, Mysore believes that the Asia-Pacific region is likely to gain a sizable share as governments push ahead with digitalisation and smart city projects, with China, India and Japan leading the trend.

Digital twins have come a long way since their inception, with advances in technologies such as CAD, IoT, cloud computing, and big data analytics making them easier to adopt and much less expensive. The initial release was a simple simulation-based event, where organizations created virtual versions of physical structures to test them under various load situations. Its main advantage is that the model can identify potential risks without actually investing in physical prototypes; Virtual models provide unlimited information about the product and its design in a variety of operating scenarios. "Over time, the use of digital twins has increased, mainly driven by the amount of data generated by the tool's assets. Sampath Kumar Venkataswamy, senior research manager for manufacturing insights at IDC Asia Pacific, said: "As long as it can be applied in real time, the insights generated by digital twins are greatly increased and always introduce a predictability factor for resolving runtime failures."

Over the past two to three years, digital twins have become increasingly popular in Indian manufacturing, utilities, oil and gas, although for most companies it is still in the R&D or proof-of-concept (PoC) stage. A Wipro spokesperson said: "Over the next three to five years, as these PoCs mature, we expect mass adoption of digital twins for the specific business needs of organizations."

Industries such as oil and gas, minerals, metals and mining, power, renewable energy, pulp and paper, and pharmaceuticals are focusing on optimizing products and processes more easily through the use of digital twins. "For example, a typical refinery could suffer millions of dollars from an unplanned shutdown and re-operation." "By investing in and implementing a digital twin, these organizations can stop production early and develop targeted production reduction plans to minimize production disruption," says Venkataswamy. Its application can be across the value chain, with organizations wanting to understand the impact of the supply chain and its impact on manufacturing operations, which in turn affects customer deliverables.

There are several examples of companies using digital twins. Cairn Oil & Gas was the first upstream company to take steps to implement a fully digital operations plan. In India, the privately held oil and gas exploration company deployed Honeywell Forging Enterprise Performance Management (EPM) software to improve efficiency while allowing workers to operate equipment remotely. The telecom industry is also a big adopter, with Finnish network equipment manufacturer Nokia introducing 5G digital design concepts to simulate 5G use cases. By using machine learning algorithms, the platform can leverage this technology to monitor and evaluate the impact of 5G implementations while providing automated recommendations. A Nokia spokesperson said: "This solution benefits several CSP (communications service provider) customers around the world."

As more and more car companies adopt digital twins, VE Commercial Vehicles, a joint venture between Volvo Group and Eicher Cars in India, adopts Dassault Systèmes' 3experience platform to develop and deliver innovative, high-quality trucks and buses in a cost-effective way for the country's growing commercial vehicle market. Similarly, Gurugram-based JBM Group, which operates in automotive, bus and electric vehicles, uses Dassault's virtual twin to develop a new generation of products, from concept design to manufacturing and even to after-sales support. Deepak Thakul, CEO of JBM (Public Transport Division), said: "Using this technology, we are able to map the entire new product development lifecycle on a single digital thread."

The technology is also good for the environment. "For most companies in India, carbon digital twins are part of an emerging field. They can help visually reduce the carbon footprint of equipment and processes by analyzing data. said Karp Praba Karan, vice president of Honeywell Connected Enterprise Engineering.

When it comes to building a digital twin, there is no standard cost, as it varies depending on the company and the complexity of the technology being developed. Costs come primarily from the infrastructure required to generate, store, and process data. Some other key components include digital dual software, IoT sensors, integration platforms, computing solutions, cloud and infrastructure, and training. Rajesh Gharpure, Executive Vice President and Global Head of Manufacturing, Energy and Industry 4.0, Larsen & Toubro Infotech (LTI), explains: "Digital twins are expensive to set up initially because of the technical components required to implement basic digital twin use cases such as Industrial IoT, connectivity (4G, 5G, 6G), cloud computing, AI and ML, and asset/entity perception, but digital twins are the future' The necessary foundation for the establishment of a service-based business model. ”

A typical PoCs or single-plant trial costs $20,50 to $6,12, depending on the size or scale of the operation and the complexity of the model. In addition, there is the cost of maintaining the digital twin and the necessary IT infrastructure. A Wipro spokesperson said the payback period is typically 10-40 months, with returns of more than 25 times in some cases. Take Piramal Glass, for example, which leverages Microsoft Microsoft's Azure IoT to optimize its manufacturing operations and build feedback loops between quality control and production teams. From the raw materials entering the furnace to the bottles coming off the conveyor belt, the company has been tracking every step of the entire production process. Shivir Chordia, head of Microsoft's Azure business group in India, said: "With Azure Digital Twins, Piramal Glass has reduced manual data collection by 5%, increased employee productivity by <>%, and reduced defects by <>%. Nikhil Malhotra, Head of Global Manufacturing Labs at Tech Mahindra, explains: "The ROI is a reduction in operating costs, increased productivity through automation, and a better operating and operating model." The IT company has created a complete training scenario for its energy customers in India, where the digital twin is not only a digital representation of the scenario, but also trains the company's employees around plant operations, thereby reducing operating costs and increasing employee productivity.

For companies, adopting this technology has several advantages, such as reduced operating costs, increased productivity due to automation, and even the discovery of some errors that would otherwise be difficult to find; However, the full potential of the technology can only be realized if challenges such as the availability of high-quality and real-time data, reliable connectivity and infrastructure, and upskilling and reskilling are addressed.