Published on 05/07/2017 | Technology
Digital twin is not a super novel concept, it has been around for years but has mostly been used for expensive equipment like jet engines.
The concept for those who are not familiar with it:
So let’s say you have a jet engine out there logging thousands of miles, it has sensors that collect all kinds of data (internal and external) based on which preventative maintenance is done, time to replace is considered, etc.
The concept of a digital twin is that while this aircraft engine is doing its normal job, the same data is replicated elsewhere, and people can perform "what if" scenarios to improve performance, schedule preventative maintenance by predicting maintenance needs, extend time between scheduled maintenance cycles, improve fuel efficiency and such based on the operating parameters and conditions. The output from these simulations is used to fine tune the performance of those engines, increase their lifespan, reduce maintenance cost over lifetime, increase energy efficiency, greatly reduce risk of unexpected failures, etc.
You can take this one step further, if the data is anonymized and correlated with a representative group of a particular category one would be able to see where your performance lies in comparison with similar profiles in the group. The data from those other profiles can also be used to suggest what adjustments could be made in order to achieve optimum results.
Different applications for the digital twin concept:
Now you can take this concept and apply it to any aspect of your business, it does not matter which industry you are in. What would have been difficult to do a few years to go is now achievable thanks to advances in sensors, big data, IoT, AI and machine learning. It would be a teaser to end this post with that statement, so let me follow it up with a very simple example:
You are a HVAC (Heating, Venting, Air Conditioning) company that maintains large HVAC installations at hospitals, malls, schools, industries, office buildings, etc. Energy savings is a significant part of performing HVAC effectively. You have things you cannot control (outside temperature, humidity, etc.) and things you can control such as the temperature settings for the building. Normally this involves quite an amount of guesswork and the human experience comes in – the knowledge of the operator who has been doing this for years. However, if you create a digital twin based on the data that is being captured – you can perform “what if” analysis and come up with optimal energy consumption levels without affecting daily operations – you can automatically set the operating parameters based on the ones that you derived from the digital twin.
Since that was a very industrial example, let us try a financial example:
You are running a manufacturing business, you have purchase contracts for the raw materials that come in, and sales contracts for the finished product you are selling. Each one of these contracts depending on the vendor/buyer has different terms of payment. Let’s say that is 30 days on both sides of the equation – this affects your bank balance. Then you have your raw material acquisition strategy, and your sales numbers – which in turn affect inventory. For most companies these are stored in an ERP system. Now if you want to fine tune the financial operating efficiency of the business – you need some smart people with experience, who can make adjustments to these contracts, payment terms, sales process, etc. and you find out the results in the next quarter if you are lucky.
WHAT IF I create a digital twin of your ERP system which still gets the live feeds of all the transactions going through but you are able to play with the terms and conditions of buy side vs. sell side, inventory management, etc. and determine the ideal financial strategy for your company to operate in based on this simulation. Your business goes on as usual until you find the answers to the questions you are wondering about and once you have those answers, you make the necessary changes for your business to realize the benefits – all this while not affecting day to day operations while you are running your “what if” scenarios.
The original article can be found here.