cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
gabrielgomane
Retired

What exactly is a digital twin, and what’s its business value?

A digital twin typically refers to a non-physical representation of a physical thing used for predictive analytics. This is most commonly known from the manufacturing space, where critical components may report usage data to their digital twin, such that the twin can be used to predict potential failures.

For example, a jet engine may report certain details of its use: hours, environment, etc. back to the twin, enabling more accurate determination (or prediction) of maintenance needs. This more accurate determination can save significant costs on routine maintenance while reducing the risk of failure.

What does enterprise architecture and digital twins have to do with each other?

Traditionally, very little. However, there is now the idea of recognizing that the models typically captured within an enterprise architecture would enable the digital twinning of the enterprise itself. In essence, a digital twin of the organization.

What’s the impact of a digital twin on enterprise architecture?

Digital twins, when applied to the enterprise, can present another opportunity of demonstrated value for the enterprise architecture practice. Digital twins can help you develop new recommendations and roadmaps, and when applied to models, they can virtually determine how different scenarios will play out. With a digital twin, you can efficiently test products, strategies, roadmaps, and models before making any real-world changes – how incredibly powerful and what a great way to be more efficient, effective, and innovative.

According to research published in Top 10 Strategic Technology Trends for 2019: A Gartner Trend Insight Report, “75% of the end-user organizations Gartner surveyed that were implementing Internet of Things (IoT) projects said they had already implemented digital twins or would implement them soon. 87% of end-user organizations implementing digital twins said they update digital twin data models as their physical assets and equipment evolve.” Technology leaders must plan for the future – and that means planning for the use of digital twins. The IT landscape is only going to become increasingly complicated and diverse, leading to increasing challenges and risks. Enterprise architects can mitigate risk and navigate complex IT issues through a digital twin’s real-time view of a company’s processes, IT portfolio, and assets.

How would a company achieve business operational excellence through an enterprise digital twin?

Using a digital twin to test scenarios is an excellent way to achieve operational excellence, as well as prioritize investments in line with expected and achieved growth. This would be achieved by using a digital twin to capture information about the business operations and present valuable analytics to the business for decision-making. In a mature case, this may include leveraging the monitored data, and running simulations utilizing the monitored data as well as data representing the trends of the actual data to determine potential revenue opportunities, or points of failure.

For example, MEGA was approached by a company that was expecting significant growth in the next few years. They were concerned with their ability to sustain their service offerings while managing the growth. The vision we had was to utilize a digital twin of the enterprise, collecting operational data related to the company growth, including indicators that were related to the concerns expressed by the company. Using this data, and the trends of that data, the company could then assess the effects of the growth on the organization and make the appropriate pivots and investments needed to sustain their vision.

 

Create a digital twin of your organization, let’s outline a three-step approach:

1. Map your organization to create a digital twin

Enterprise Architecture helps you create a digital twin of your organization. It provides you with a detailed understanding of your organization by modelling each piece of it: strategy, business capabilities, processes, customer experience, data, applications, and infrastructure. All these elements are tied to one another in a single platform, so that you can easily perform impact analysis and help transform your business. It also allows you to ensure that strategy is well executed at a minimum cost while improving time-to-market. But all these models are not alive! To make them alive, you need to infuse some blood into them with real-life data. Real-life data will help you optimize your models by giving visibility into what really happens. But to do so, you’ll have to define KPIs first, so that you can reduce the scope of what you need to measure.

2. Identify KPIs to narrow down the scope

In this step, define measurements based on your business objectives or other factors. By doing so, you can limit the amount of collected data and only focus on the ones that are of interest to you. For example, you can monitor the lifecycles of the software technologies that underlie your business applications, putting potentially at risk business capabilities that use these applications. You can measure customer satisfaction on the various touchpoints of a customer journey. For a specific process, you may also want to measure the time to execute a task, or check if an order has not been paid twice in SAP for example. KPIs can also be defined more broadly at an enterprise level, such as revenue growth, customer satisfaction or EBITDA.

 

3. Perform and analyze data mining

So far, you have created models and identified KPIs for your organization, now you can improve its efficiency by incorporating real-life data. This requires analyzing event logs stemming from information systems such as ERPs or CRMs to identify trends and patterns. Use APIs to connect to these systems and import the data relevant to your analysis. This analysis is performed on a continuous basis to discover inefficiencies, to check whether real-life data conforms with what have been modeled, or simply to discover undocumented processes in the organization.

For example, by analyzing real-life data, you can realize that the same task can take twice as long in one branch as in another. You can then correct your processes and ensure there won’t be any further deviations to the process model that was initially defined. As another example, by examining patient records in a hospital, you can identify new processes that have not yet been documented and hence, improve patient management. You can also fuel satisfaction ratings at the various touchpoints of a customer journey based on customer feedback systems and provide recommendations on how underlying processes can be improved.

 

In summary, to create a digital twin of your organization, create business and IT models as a starting point. Then, define KPIs to narrow down the scope of measurements. Finally, based on a continuous analysis of real-life data, optimize existing processes and EA models. By performing these steps, you get a clearer view of your organization so that you can efficiently tackle business transformation challenges with more accurate models.

 

How do you see digital twins impacting business in the next 5 years?

I expect to see an increase in attention put on digital twins of the enterprise in the next 5 years. There will be a maturing of understanding what data needs to be collected and tested in order to derive valuable analytics. Additionally, the ease at which this data can be collected and managed will continue to increase with the increased adoption of technological sensors. As this maturing of both understanding the potential, and the ease at which data can be collected (or performance measured), we expect to see increased opportunities to expand on the potential benefit of digital twins of the enterprise.