Services
IoT and Digital Engineering
Bringing Together IoT and Digital Engineering to Create Rich Experiences
Highlights
- Together, digital thread, generative AI, and knowledge graphs allow deep discovery of vast amounts of data to generate new intelligence and newer contexts.
- They provide deep insights into the complexities around engineering data and can open new avenues to rapidly generate new-age product designs.
- When combined with knowledge graphs and generative AI (GenAI), the business transformative potential of a digital thread goes up exponentially.
- Companies can unlock greater value from product life cycle data, enhance traceability across the end-to-end product life cycle, and drive their product life cycle strategies.
On this page
On this pageinpage
Digital thread
A digital thread establishes digital continuity across the enterprise value chain, helping overcome process and system disconnects.
Manufacturing companies are battling multiple challenges—new business models, the pressure to deliver sustainable solutions and outstanding customer experiences, evolving ecosystems, and increased product complexities. They are struggling to quickly roll out products, be more responsive to change, control costs, and improve productivity and operational excellence.
Overcoming these challenges will require them to simplify and integrate their disconnected business processes and systems and accelerate their digital transformation. Building a digital thread to establish digital continuity across the enterprise value chain can minimize the process disconnects and bring traceability to product data and processes. This will give companies control and enable a seamless flow of contextual product data required to make timely and informed decisions.
Digital threads are the data foundations that power digital twins and are continuously updated with insights and recommendations from these twins. They are characterized by four key behaviors:
- Model-driven: Product life cycle artifacts and data are built as models (as multi-computer-aided design models, simulation models, etc.) that provide access to various product characteristics and their effectiveness.
- Connectivity-focused: Establishing the interconnect between these artifacts and data as a product evolves from the concept stage to a fully operating asset in the field and beyond is critical to enhance offerings and drive greater business value. So is realizing seamless connectivity between engineering, manufacturing, and service functions and their respective suppliers.
- Insights-driven: Digital threads typically involve a mechanism for accessing and harnessing the artifacts and data to generate intelligent insights (service intelligence to enrich customer experience or performance insights of products and product life cycle processes to drive better innovation), predict future states, and make proactive intelligent decisions.
- Closed-loop actions: Digital threads allow companies to implement decisions based on insights from functions such as engineering, manufacturing, field service, and supply chain, thereby closing the loop generated between these functions. A closed-loop engineering change intelligence covering insights from engineering, manufacturing, and field service can, for instance, drive customer responsiveness.
The power of three
When combined with knowledge graphs and generative AI (GenAI), the business transformative potential of a digital thread goes up exponentially.
Together, they allow unprecedented levels of data discovery on vast amounts of data to generate intelligence and new contexts. This approach is perfect for handling product life cycle data.
A key challenge that the industry has been grappling with is modeling product life cycle data using industry-standard ontologies and trying to achieve seamless interoperability between different stakeholders who work on diversified systems. The diverse nature of data and data formats makes the task daunting. Knowledge graphs are well-suited to modeling and delivering a highly networked web of data. This makes them a perfect complement to digital threads, which can be visualized as a web of complex, diversified product artifacts and data covering the entire life cycle of a product. Together, they can help companies unlock greater value from product life cycle data, better visualize traceability across the end-to-end product life cycle, and drive their product life cycle strategies. They can:
- Provide a scalable, intelligent, and interoperable platform that can easily support AI algorithms, semantic queries, large language models (LLMs), and GenAI solutions.
- Be easily integrated with low code, no code platforms for persona-based and contextual visualization of a digital thread.
- Simplify and run queries on complex networked data to generate contextual information. Domain-specific graphs that scope the boundary of data in context can make it even easier to query and generate insights.
- Help fine-tune LLMs that organizations could use to generate specific outcomes from GenAI. Knowledge graphs’ ability to handle structured, semi-structured, and unstructured data easily can help prepare the data for this and multiple analytical purposes.
- Support industry-standard protocols like resource descriptive format for data formatting and exchange, and industry-standard ontologies like AP242, which are available in graph formats.
Persona-based
There are many exciting real-world applications for various stakeholders in product development.
Different personas involved in product development need different contexts of product life cycle data for their day-to-day decision making. With a knowledge graph implemented on top of an enterprise digital thread and used as a platform for running AI and GenAI algorithms, persona-based views can be easily generated for various business contexts. For instance:
- Parts engineers can get recommendations for the reuse of existing parts and designs when developing new products. This can drive parts reuse and reduce costs.
- Design engineers can generate designs based on data from knowledge graphs and model-based systems engineering practices. By doing so, they can promote design for sustainability.
- Engineering change analysts can conduct a comprehensive engineering change impact assessment to obtain a complete picture of the impacted artifacts and data. This can increase change accuracy.
- The engineering change review board can get a full picture of the risks involved with different potential solutions, allowing them to choose for implementation the most appropriate and cost-effective solution with minimal potential risk and the least implementation cycle time.
- Compliance engineers can quickly generate a dossier of various documents to be submitted for compliance purposes, saving on the time taken for approvals.
- Specification engineers can generate a specification or proposal response based on customer requirements, making the organization more responsive to customer needs and accelerating the response time.
- Managers and executives can get an intelligent view of the complete traceability of the product life cycle, the field performance under real operating conditions, failure predictions, and impacts on manufacturing and engineering.
Knowledge graphs, digital threads, and most recently GenAI have gained traction in the industry and are increasingly being used to build scalable, intelligent, and interoperable solutions.
Architecture
To fully exploit the combined potential of knowledge graphs, digital thread, and GenAI, enterprises need a resilient, scalable, and interoperable architecture.
Such an architecture can help deliver innovative solutions to address data complexities, allow better visualization, and enable intelligent analytics. We envision a multitier architecture integrated with enterprise systems and comprising five key layers:
- Digital thread layer: This serves as a data information map, pulling in data from heterogeneous data sources and representing various product life cycle functions.
- Ontology model layer: Industry-standard ontologies used to represent and model the complete product life cycle can be easily leveraged to provide standard representation and seamless interoperability between diverse systems.
- Knowledge graph layer: Data from multiple systems and ontology models are integrated to form a knowledge network. AI/ML algorithms are used to automate the creation and management of this network in an intelligent manner.
- Vector search layer: Product data available in structured, semi-structured, and unstructured forms can be easily modelled in a vectorized format and used to fine-tune LLMs with organization-specific data. This approach allows searches for identifying similarities and anomalies in product data drawn from various complex, legacy, and duplicate systems.
- GenAI/ML operations services layer: This helps build various services to manage large data sets and convert data into vector form. Different types of GenAI-powered assistants and bots can be built to automate processes and generate new contexts for different business scenarios.
An architecture for knowledge graphs, digital thread, and GenAI
Use cases
The power of three—knowledge graphs, digital thread, and GenAI—can be effectively used to reimagine problem-solving in product development.
It can be applied across industries, as shown by the high-impact, cross-industry use cases below:
- Managing product complexities in manufacturing: Products are becoming feature-rich and more complex thanks to the software-driven nature of products today. With knowledge graphs implemented on digital threads, companies can easily model, manage, and visualize the product and product life cycle data complexities, and perform more accurate design and architecture trade-off analyses.
- Healthcare regulatory submissions: The highly regulated healthcare industry mandates a very stringent product approval process involving submissions of various artifacts with comprehensive detailing of product characteristics for compliance and certification. Generating and submitting this documentation is a mostly manual and time-consuming affair that’s also error-prone. Leveraging GenAI assistants, users can quickly generate these documents, combining past data, templates, standard protocols, and new requirements, and accelerating the time taken to achieve regulatory compliance.
- Accelerating inquiry-to-order response in the engineer-to-order industry: Knowledge graphs can significantly cut down the order execution cycle time in this industry, which is characterized by a high degree of product customization and requires very quick response to inquiries. Companies can use them to quickly scan through a networked data of historical orders or inquiries of similar kinds, find the nearest delivered order, blend the information with unique propositions, and quickly generate a new order and its related engineering information.
- Improving contract manufacturing control in high tech: Products today can extensively leverage a large ecosystem that goes beyond the four walls of a company. Knowledge graphs can be leveraged to provide a role-based secure access to product IP and enable data exchange with various stakeholders.
- Assisting product specification generation in the consumer products and goods industry: Product specification engineers need to churn through various standards and regional regulatory requirements when generating product specifications. Speed and accuracy are critical due to the high variability of product portfolio. Using insights from knowledge graphs and digital threads, first-cut product specifications can be quickly generated with higher accuracy, significantly cutting review and signoff time and increasing the speed of release.
- Minimizing cost of parts across industries: Reducing the cost of maintaining a part in a product is a critical KPI to keep overall product cost under control. AI/ML services run on knowledge graphs can be used to identify closely matching parts and designs, which can be recommended for reuse, minimizing the costs of parts.
- Engineering change management: Products in the field will undergo changes to solve defects, optimize performance, or for innovative uplifts. Exercising the right engineering change controls calls for comprehensive impact and risk assessments to improve cycle time and accuracy and minimize the recurrence of the change. Knowledge graphs can be used to build a network of product data, all the life cycle artifacts, and their complex interrelationships which can be used to establish the impacts and potential risks.
- Sustainable designs: Computer-aided design and simulation models can be fed more accurate inputs from the graphs to generate optimized designs that can contribute to building sustainable products. Also, information about hazardous substances can be provided at the design stage to prevent or restrict their usage.
New paths to value
Digital threads, knowledge graphs, and GenAI can unearth impactful data points entangled in complex product data networks and unlock exponential value for companies.
Working in concert, they can boost productivity and responsiveness for companies. They can help enterprises make informed decisions, deliver cost-effective products, minimize asset downtime, or create sustainable, innovative designs that reduce friction in the customer value chain. Indeed, the opportunities for transformation across the value chain are unprecedented. Those who move in early to capitalize on these opportunities and harness these powerful technologies to identify disconnects across the enterprise, navigate product data complexities, build traceable product life cycles, and rapidly generate intuitive designs are already on the path to greater success.
