Digital Twin For smarter, safer roads

The NASA website will tell you that the idea of a Digital Twin (DT) was born in the 1960s as a living model of the Apollo mission. In response to Apollo 13’s oxygen tank explosion and subsequent damage to the main engine, NASA employed multiple simulators to evaluate the failure and extended a physical model of the vehicle to include digital components. This digital twin was the first of its kind, allowing for a continuous ingestion of data to model the events leading to up to the accident for forensic analysis and exploration of next steps… While elaborating on DT, Rajesh Krishnan, CEO, ITS Planners & Engineers Private Limited, points out that the DT technology holds a great potential for the future in transport and other domains.

A DT creates a digital replica that serves as an identical counterpart to a specific physical component (a valve or a motor), an object (a vehicle or a building), a system (power grid, road network) or a process (manufacturing, supply chain). DT can be used for a wide ranging applications such as designing, process optimisations, what-if scenario planning, preventive maintenance and as a training tool.

We need to adopt a process driven metric rather than a people driven one. Decisions must be data driven with clear objectives of what you want to achieve. That context within which digital twins are deployed is important, without which the technology will be just another novelty which will fall by the wayside when a new fad takes over.”

DT technology is becoming increasingly mature in several domains such as manufacturing, supply chain and logistics, automotive, construction, healthcare and retail and its capability is rapidly evolving. Are we putting this technology to public use? Could we have reduced the devastating impact of the recent cloudburst at Uttarakhand had we seen the scenario in advance? Can we achieve our aim of Net Zero deaths and reduce road fatalities by 50% in the near future with our current resources? Can we reduce congestion and pollution in our cities using DT technologies?

The answer is, as with many questions of this nature, depends! DT means different things to different people. A DT could be a detailed 3D model of components or an object whose data gets updated based on measurements made by IoT sensors. Such a DT would be suitable for easy understanding of the real-time state of a system even for people
who are not familiar with the domain, but it would have no predictive power.

On the other hand, process DTs would be capable of predicting the evolution of a system over time based on currently observed data but may come with abstract visual representation of the process state that could only be understood by domain experts. Ideally, we would want the best of both worlds with a DT having advanced visualization and predictive power. Before we delve into this further, let us demystify some of the terminology around DT.

We all agree that DT is a model with a connection to its real-world counterpart – without the connection, it is called a Digital Model (DM). In the transport domain, a micro-simulation model is a DM. There are models that are automatically updated based on real-time data. In some of the academic literature, these are not DTs – they are called Digital Shadows (DS). They shadow the state of its real-world counterpart in near-real-time. DS can be used for analysis; Hawk Eye used by third umpires in cricket for determining whether a ball would have hit the stumps is a DS.

 

DT on the other hand should have two-directional data flow. It is a DS where the actions based on inference from the DS gets implemented in the real-world – either manually or automatically. However, in some contexts, DTs where the real-world is updated manually using information from the DT is still considered DS. On the other hand, in many contexts, the difference between DS and DT is blurred and both are called Digital Twins.

The demystification of jargon is not complete unless we discuss related acronyms that are merrily bandied about such as Building Information Modelling (BIM), Cyber–Physical Systems (CPS). BIM is a collaborative approach for creating, managing and visualizing three-dimensional digital models of buildings and projects, and the data flow is typically unidirectional in nature.

 In the area of transport, Area Traffic Control Systems that carry out model-based optimisation are DTs as per this definition; the SCOOT traffic control system deployed since early 1980s use DT! However, there is an expectation of advanced visualizations, if not 3D visualizations, when it comes to DT. Modern ATCS systems provide estimation and prediction of traffic states graphically can be categorized as DT.

It is necessary to differentiate DT from related concepts such as building BIM, CPS and digital shadows, according to the research material. DT emphasizes the bidirectional data connection between a physical entity and its corresponding virtual model, whereas BIM lacks this real-time connection and is merely considered a foundation for DT construction.

A CPS refers to an integrated system wherein physical components such as sensors, actuators and machinery interact with digital components, including software, algorithms and networks. In CPS, the mapping between physical and digital components is one-to-many, while in DT, it is strictly one-to-one.

Can DT technology be used to predict natural disasters like the cloudburst event that recently happened in Uttarakhand? We understand the science behind climate, and we can model climate events provided we have the data. Weather related data collected from satellites along with on-ground weather data such as temperature, humidity, pressure, wind etc. along with historic weather data and information about past cloudburst events can indeed be used to model and predict the probability of cloudburst events in the future. One needs to ensure that on-ground sensor coverage provides the required data for such modelling and DT creation.

Transport DT models are already deployed in many of our Smart Cities in India. Such DT models use real-time traffic flow data from sensors typically installed for ATCS to estimate the real-time traffic state of the road network. In other words, the number of vehicles on links, their destination and speed. The DT can be used for making traffic control decisions and optimising traffic signal timings. Such DTs also estimate traffic on the entire modelled road network, including the links that don’t have sensors, using traffic flows from a smaller number of links. Hence DTs can be used to reduce the cost of sensor deployment.

 The use of models is more as a reference and they get fed into this framework. One of the things a DT does is interpolation of what happens between measurement points. So, you do actually get more benefits for the money that you invest if you add a digital twin.

For e.g. if you want to know what is happening on this road, you don’t need to add a sensor on that road – you can have sensors distributed strategically and the digital twin will interpolate. That is an economic Use Case that people sometimes overlook because DT will oftentimes have a process model built in, so from partial observations you can estimate what goes on in the whole system. Traffic DT can also predict the impact of planned events such as roadworks and hence act as a decision support tool.

Quoting Peter Drucker, unless you measure something you cannot improve it. When you translate that in the context of digital twin, unless you model something you cannot optimise it, especially when it comes to traffic and transport using real time data. Digital twins are very important in helping to make data driven decisions. A lot of the operational decisions that people make are based on gut-feel and experience. We need to move away from that and follow a scientific approach, and DT technology can help a long way in this regard.

While we must encourage the use of DT, we cannot realise their full potential unless we have people who have the required technical knowledge about both in their domain and about the DT technology. While the DT technology holds a great potential for the future in transport and other domains, we should not overlook the training and upskilling of professionals in order to realise its full potential.