Digital twins technology with IoT

 What is digital twin?

Digital twin is increasingly becoming popular since 2018, as the virtual replicas of physical assets. Simply said, this buzz word refers to a technology that helps carry out features like device simulation during development, ingestion of real-world data about a physical object or system as inputs and producing the outputs or simulations based on those inputs helping scientists and IT professionals run simulations before actual devices are built and deployed.

Digital twin technology has now moved to multiple industries and vastly merging in the Internet of Things, artificial intelligence and data analytics helping augment deployments for peak efficiency and create other what-if scenarios.

Via simulation of real object and its interactions with its surroundings, this technology helps provide a more accurate representation of the shape an object than a physical replica.

The power of digital twins can be extended to virtually any technology such as cloud computing, artificial intelligence, machine learning etc.

While not entirely new, but this concept of creating twins to help better decision-making has long been existent such as factors around computer vision, artificial intelligence, machine learning, and advanced simulation

How does a digital twin work?

A digital twin model is constructed with the ability to receive input from sensors gathering data from a real-world physical counterpart allowing near realistic simulation in real time, helping gain insights for performance and potential problems. This helps providing feedback as the product is refined. Alternatively, a digital twin can also even serve as a prototype itself before any physical version is built.

In order to create a digital twin of a physical asset, the technical team collects synthesized data from a variety of sources, including sensors and physical components of assets such as buildings, cars, vehicles, equipment’s and other physical objects.

This data is then processed with AI algorithms to translate into a virtual model with applied analytics on top for optimum analysis and insights about the asset. Supported by consistent flow of data helps fine tune business results.

While largely still in infancy stages, the best way to build digital replicas is by attaching structural sensors acting as boundaries to the digital platform in terms of replicating shape, and object.

What are the types of digital twin models?

1.      Digital twin prototype (DTP)

2.      The digital twin instance (DTI) and

3.      The digital twin aggregate (DTA).

The DTP prototype model is made up of designs, analysis and the overall processes needed to manifest a physical product model ahead.

The DTI on other hand is focused more towards the digital twin of each individual instances of the overall product once it is manufactured.

The DTA is the cumulative build of all DTIs whose data and information can be used for overall analysis about the physical product, its related prognostics, and learning driven by use cases.

Image source- https://internetofbusiness.com/half-of-businesses-with-iot-projects-planning-to-use-digital-twin/

Digital-twin use cases

If we analyze the use cases, such technologies can be used for aircraft engines, offshore platforms among others are designed and tested digitally before being physically manufactured.

In a IT related change management approach, these digital twins could also be used to help with maintenance operations as a means of digital twin model being subjected to proposed fixes before applying the fix on the physical twin.

Digital twins and IOT

Digital twins can be leveraged to predict different outcomes based on data. This can now be also used with additional software and analytics, IoT deployments can be done for maximum efficiency, as well as help designers figure out where things should go or how they operate before being physically deployed.

Digital twin vs. predictive twin

Digital twin is all about 3D rendering and details on all the sensors in the device continuously generating sensor readings simulating real life options.

Predictive twin on the other hand is about modelling the future state and behavior of the device based on historical data from other devices, simulating breakdowns and statistical analysis.

Cyber and twinning concept:

The digital twin concept also blends in well with cyber systems allowing integration of physical and virtual systems into cyber security model.

Cyber-Physical Production Systems also on other hand helps smart factories to assist in various decision-making processes by predicting the future based on evident past and analysis of present situations.

A major benefit which can be drawn upon is since the digital twin runs in an isolated virtual environment, it helps analyze without the risk of affective live systems

Risks in digital twin technology?

1.      Inaccurate representation of an object using digital twin

The biggest concern the consumers of this technology would have is the possibility of improper representation of this object or system being replicated as not sufficient research has been done yet to validate accuracy of digital twin models compared to its physical counterpart.

2.      Replicating inside anatomy of models:

Based on the inner working complexity of certain models, it might be challenging at times to replicate the inside portion of your object. This is still a matter of debate but to have maximum precision, the digital twin software would need to be augmented with performing manual revisions in order to ensure the accuracy of the simulation.

3.      The accuracy of futuristic simulation using digital twin

Two factors which need to be counted in are the level of accuracy of the simulations and rules appended to the amount of data the digital twin has accumulated from the physical twin.

4.      Affordability of utilizing digital twin technology for small businesses

The technology adoption and deployment costs are another factor to be considered for flexibility on every company’s budget, so not only the big players in market, even the smaller ones can harness the benefits of this technology

As with all risks in technology, there is also a silver lining in cloud with respect to advantages, digital twin technology can be used to vastly improve posture of cyber security. A simulated cyberattack can be detected by the digital twin using virtual databases to capture information and testing activities which can be used to create cyber algorithms to defend the firm’s data against malicious viruses.

Digital-twin vendors

·        GE has developed digital-twin technology internally as part of its jet-engine manufacturing process

·        IBM is marketing digital twins as part of its IoT push

·        Microsoft is offering its own digital-twin platform under the Azure umbrella.

Microsoft

Microsoft Azure IoT has the concept of a ‘device twin’ , This is a model which gets automatically created when a device is connected to the MS IoT Hub. The device twin gets created as a JSON file which has the capacity to store device state information to have synergy with back-end processes.

Amazon

Amazon has the concept of  ‘device shadow’. Similar to Microsoft as a JSON file, it gets created covering state information, meta-data, timestamp, unique client tokens and version of a device connected to the device shadow service. 3 basic REST APIs that can be used to interact with the device shadow are: GET, UPDATE, DELETE. There’ also option communicate with device shadows using MQTT messages.

Reference: https://dzone.com/articles/the-reality-of-digital-twins-for-iot