How digital twins can make an organisation alive and intelligent using automation

According to a 2019 study from Mindtree, 77% of organisations have implemented some artificial intelligence-related technologies in the workplace. These artificial intelligence technologies are already driving tangible value, but most often address only specific segments of the business, making decisions or guiding human decisions in limited areas. Once we combine artificial intelligence with the broader concept of a digital twin of the organisation, businesses can automate decision-making not just in disconnected parts of the company or processes, but across the enterprise. 

A digital twin is simply a digital representation of a company and its processes. The combination of artificial intelligence with digital twins will deliver a new modality, intelligent process automation, as executive teams apply artificial intelligence to solve broader problems in their business. 

What role do digital twins play in this progression towards enterprise-wide artificial intelligence?

An artificial intelligence algorithm used to optimise a narrow slice of the business must have access to and understand data about the specific business function involved in that specific area. In order to deploy artificial intelligence across the broader business, that algorithm must be able to understand and access information and metadata explaining the importance of that information through the entire business value chain.

A digital twin captures all areas of your organisation including the strategic, tactical and operational relationships between different divisions, functional units, equipment objects, assets or personnel roles. An artificial intelligence algorithm, or series of algorithms packaged as reusable cognitive services, can use the digital twin to understand and solve different business problems, and truly automate the entire organisation.

The digital twin gives you the tools to facilitate artificial intelligence across your organisation, handing entire business functions over to intelligent process automation-driven decision support and automated decision making. This is not an entirely new concept, and enterprise software companies have been moving in this direction for some time by developing operational intelligence, tools for business. 

Operational intelligence provides executives with real-time and dynamic business analytics that yield insights not just on static data, but streaming events and business operations. This enables them to move through the observe, orient, decide and act, OODA loop faster than their competitors.

An artificial intelligence-powered digital twin as we have described it is the evolution of classic operational intelligence software and gives executives real-time and actionable operational dashboards and support for making decisions on a course of action. In aviation, drone aircraft are the latest step in a natural process as the entire observe, OODA loop is automated. 

An artificial intelligence algorithm can look at and comprehend infinitely more data points and dynamics than a human pilot and make optimal decisions in the moment. Similarly, in business, a digital twin can support executive decisions and lets an executive team automate the execution of these decisions. 

The algorithm will make decisions accordingly, more efficiently than any human mind ever could. Digital twins will extend this approach to the business strategy level, enabling an executive team to ask algorithms to optimise business operations for short term revenue, customer retention, new business acquisition or other enterprise-wide goals. 

There are four areas that help bridge the gap between decision support and automated decision-making in a digital twin context.

#1 Knowledge representation

Knowledge representation means classifying and cataloging unstructured data: cognitive computing techniques like cognitive search and text analysis can make sense of this data and to build it into an intelligent and accessible knowledge base. 

First-time fix rate in field service repair will greatly benefit from an intelligent work order that can automatically parse, interpret and present the information a field technician will need to complete a service call, from manuals and documents to previous resolutions and lists of necessary parts.

#2 Intelligent decisions

Artificial intelligence and machine learning can then add context to these internal and external data sources and monitor processes, learning from their execution. Artificial intelligence-powered enterprise software will predict deviations from the normal or optimal flow and even recommend improvements. The combination of advanced support to decision-making and automation is what ultimately will lead to truly intelligent decisions. 

It is important to point out, though, that artificial intelligence used to support business processes must be as explainable as possible to help users understand the rationale behind decisions and recommendations and ultimately gain their trust over time.

#3 Autonomous execution

Machine learning enables software to identify values and thresholds for specific decision points based on past data, entirely automating decision making when possible, facilitating faster decision execution and allowing for the collection of feedback on the effectiveness of the flows. This means intelligent process automation-enabled software can continually monitor, optimise and improve the workflows themselves.

#4 Enhanced assistance 

Multiple channels can be used to allow users to interact with intelligent knowledge bases and to drive value from the automated processes. Conversational user interfaces are already a reality for enterprise software and enhanced assistance is offered to office and field workers alike via chatbots and even through augmented reality channels, where information from the artificial intelligence-supported knowledge base can be visually presented to the field worker.

Intelligent process automation will probably come to market as a built-in feature enterprise business software as early as this year. The time to start preparing mentally for this eventuality is now. Organisations that creatively think ahead, envision the potential and make plans to manage the necessary changes will be the first to leverage intelligent process automation for competitive advantage. 

Key takeaways 

• Drone aircraft are the latest step in a natural process as the observe, orient, decide and act loop is automated. 
• A digital twin is simply a digital representation of a company and its processes.
• An artificial intelligence algorithm must have access to data about the specific business function involved in that specific area. 
• Algorithms must be able to understand and access information and metadata explaining the importance of that information.
• An artificial intelligence algorithm can look at and comprehend infinitely more data points than a human pilot.
• Digital twins enable an executive team to ask algorithms to optimise business operations.