Getting started on your first digital twin solution
Updated: Sep 16, 2020
What exactly is a digital twin?
Gartner defines a digital twin as “a digital representation of a real-world entity or system.” However, some key concepts are missing from this definition. A 3-D rendered model also provides a digital representation of a real-world entity or system, but it remains static and focused on the entity or system's geometry, layout, footprint and other spatial properties.
A digital twin, on the other hand, collects near or real-time data from an array of IoT sensors, online devices and connected assets. And once it has that data, it can analyze the information to improve various aspects of performance.
I think of digital twins as "a virtual or digital replica of a physical asset(s) or processes, with the ability to monitor, optimize, improve and control operational and business processes, as well as simulate scenarios based on varying input parameters".
Why are digital twins relevant?
In volatile and high investment environments such as oil and gas, mining, defence, manufacturing and transportation sectors, investment in digital twin capabilities can be relatively inexpensive compared to the entire asset or business unit shutdowns, catastrophic component failures and disastrous return on investment levels.
At the start of 2019, according to Gartner research, 24% of organizations had either IoT technologies in production already, or IoT projects in progress that made use of digital twins. The same study found that another 42% plan to employ digital twin solutions within the next 3 years.
The potential benefits go beyond traditional cost reduction and efficiency efforts and tap into the hardcore "where the rubber hits the road" kind of real-time involvement. If you ever wanted to get closer to the source of understanding and addressing a problem, a digital twin is an excellent way to do that.
Where do you start?
There is no single correct way to follow from start to finish; each instance and solution is very unique, but in our experience, there are basic principles that must be followed to ensure success. Such a major and complex initiative inevitably impacts 3 core areas of the business, namely (1) people & culture, (2) processes & practices and (3) tools & technology.
Below I have listed 15 aspects that, if considered, will help guide and keep your efforts on track. The points discussed each extend into a far greater level of detail which will individually be addressed in future posts. For the purpose of this post, they are kept light and purposefully brief.
People and Culture
(1) Have a solid value proposition
We have all heard the saying: Avoid technology for technology’s sake. This may even be more relevant today, given availability and affordability of technology solutions. Just because a business can adopt technology, does not mean that they have to.
Adopting technology can and will make lasting changes in your business, so backtracking is not an 'easy-out'. Quantify and qualify the vision, value and return before making an investment decision. Articulate the exact problem that is being solved.
(2) Drive ownership and accountability
Put someone in charge who is passionate and determined, regardless of whether they are in IT.
"Most people have a very strong send of organizational ownership, but I think what people have to own is an innovation agenda, and everything is shared in terms of the implementation" - Satya Natella, CEO Microsoft
The role and involvement of the IT department, and the shills, capability and vision they bring to the table is very necessary and critical. But just because the solution is digital, and technology-driven by nature does not mean that it must be head up by them.
Pick a visionary, a dreamer, a big picture thinker and provide them with the necessary tactical and technical resources and skills to be successful.
(3) Seek out and develop new skills
Look for agnostic or skill transfer opportunities; by doing this you will gain a deep and sincere appreciation from the existing workforce. Being involved with such a huge transformational project is often the dream of many tradesman, admin staff, technically inclined individuals, non-operational folks and finance professionals.
Digital twin solutions and true transformation often thrive and become enriched with diverse input and collating a range of vision and perspectives. With such a huge undertaking the opportunity would be wasted to limit yourself to the same thinking that has dominated the landscape.
And be creative to attract new skills. The future will look very different than what we know it to be now.
(4) Communicate, share, repeat
Do not keep stakeholders in the dark. Communication is key; it is the glue that binds the dynamics between the people, process and tools together. Designing and implementing a digital twin can touch so many different parts and functions of the business, those robust communication methods are needed; beyond the normal practices already employed.
Adopt a change leadership ‘lifestyle’ on this journey and lead with transparency. Agile methods and short sprints, accompanied by demonstrations, practical sessions and working groups all form part of sharing and communicating as the process unfolds. You will not get much bang for your buck when using a conventional "communicate only for periodic updates" approach.
(5) Leadership rise to the occasion
Nothing will cause the death of your digital journey faster than lack of leadership buy-in & support. Conversely, if the value is qualified and quantified, very few executives will steam roller the initiative to be a non-starter.
It must also be said that while the very essence of digital twin initiatives relies on visioning, innovation and a forward-looking approach, the reality is that investment decisions are largely driven by dollars and cents - they do matter.
But refer back to aspect no.1 - have a solid value proposition.
Process and Practice
(6) Get a partner and collaborate:
A well-placed partner brings experience, innovation and problem solving to the table. There is no shortage or lack of players and solution providers in this space. many with untainted and solid reputations, with the weight of a global multinational behind them like GE, Aveva, ABB, Schneider, SAP, Oracle, Microsoft, IBM, Siemens, etc.
Various studies have shown that the IoT and digital twin market share is expected to witness somewhere between 30 to 35 percent growth by 2025. This means that some serious players are going to make a serious investment into developing the solutions and technology so that you don't have to do it yourself.
(7) Clearly map out the performance drivers
Align the performance drivers with the vision. Be clear on questions such as what will be measured, the measurement intervals, a fit for purpose response matrix, the decision-making process and protocols, etc. Then use the technology and solutions on offer to gather the right data, to convert it into the right information, to get it in front of the right people at the right time.
In itself this aspect is unique, can be complex and is intimately integrated into the design and implementation of the digital twin. As mentioned at the start, I plan to expand and dive deeper each aspect in the future.
(8) Pay meticulous attention to detail
Digital solutions bring the ability to manage terabytes of data in real-time, so leverage that capability to your benefit. The life and energy can easily be sucked from the team and momentum can quickly be stalled when the solution brings more work to the table due to unorganized data.
Upfront involvement of the necessary stakeholders, clear data gathering and analysis requirements, adequate ‘sensorizing’ and making infrastructure design and performance are ways to help solve the problem.
(9) Follow the start, stop and continue rule
Understand upfront the processes and practices that will be either introduced (start), made redundant or outdated (stop) or remain in place (continue).
With the magnitude of data and analytics that is involved with digital twins, there needs to be clear processes, with clear roles and responsibilities, for activities associated with infrastructure reliability and maintenance, data gathering and cleansing, data storage and documenting, governance and control, reporting and monitoring and governance.
Doing a business-wide purge will prevent frustration and usher in a new era of lean and focused mindsets.
(10) Data and information is very valuable
There are many articles and views about big data and data management; but for me, the easiest and most impactful to remember are the 5-v's:
Volume: the name big data in itself signifies managing terabytes of data coming at you at lightning speeds.
Velocity: the rate at which data is collected and needs to be organized, analyzed, processed for decisions and report.
Variety or Variance: the differences that exist in the data itself when being collected, often from multiple different sources and devices, and then being structured, semi-structured or unstructured.
Veracity: the degree of inconsistencies and inaccuracies that may exist within data sets, as well as missing data - in other words, the data quality or completeness.
Value: the role that data plays in satisfying the objective or solving the problems that the digital twin was initiated to do.
Information and intelligence deduced from data is only as good as the input data used as building blocks. Garbage In = Garbage Out.
Tools and Technology:
(11) Invest in system reliability and uptime
This will allow you to stay ahead of the curve and let you focus your valuable resources and time on innovation and value creation. It can be extremely disruptive and damaging to the success and benefits of a digital twin when the biggest amount of time and effort is spent on infrastructure maintenance and troubleshooting due to design flaws and shortcomings.
Often the analogy is used of driving your shiny new high-performance sports car on a badly maintained road, full of potholes and twists versus a smooth highway, the autobahn if you will. Your shiny new sports car will not perform as designed, neither will you get out of it what you expected and want on a pothole-riddled unmaintained road.
(12) Design discrete measurement points
Digital sensors have the ability to be installed just about anywhere and measure just about anything, so make it count. I already mentioned big data and incorporating the expectations into the design process as early as possible can save valuable time, effort and costly rework solutions later.
Knowing the location and very specific function of each sensor and each kilobyte of data to be gathered will not only be a huge factor for progress and ultimately success but can potentially also drive down costs significantly.
(13) Put system integration and scalability front and centre
At the rate that technology is advancing today, a business cannot afford anymore to be saddled with massive systems that are not flexible and are cumbersome to operate and maintain.
Modern tools and apps, as well as a whole new range of available skillsets, can now allow for in-house changes and re-programming in order to adapt to changing business needs and operating environments.
When the way that the world will work in the future is unknown, you cannot design for it. But you can design for a system to be able to accommodate change at minimum disruption, no delays and low cost.
(14) Keep solutions accessible and user friendly
Design with end-users in mind. After all, humans will be pivotal in interacting and extracting data and information from the digital twin. The success of mobile applications on our smartphones can largely be attributed to its simplistic nature, specific purpose and ease of use. Add to that the responsiveness of design teams and programmers to end-user feedback, and you have a recipe for success.
The system should work for the people, and not the other way around.
(15) Use the full potential of cloud and mobile
Advances in technology have once again provided the gateway into cloud computing and distributed or remote infrastructure and devices. Some of the key benefits are risk distribution, cost efficiency and system performance.
Employing edge computing, smart sensors and mobile functionality the digital twin enables fast and efficient service that drives dynamic decision making. There is no need for "stop-wait-discuss-decide-and-go" based decisions anymore.
While this post seems like somewhat of a mouthful, the reality is that the surface has barely been scratched on each of these aspects. The final message that I want to leave you with is that digital twin solutions are not the daunting and overwhelming endeavours it used to be, but at the same time realizing it does need very specific focus and attention to be successful.