Digital Twins for Smart Cities
Image Source: sutlafk/Stock.adobe.com
By Adam Kimmel for Mouser Electronics
Published August 23, 2022
Digital twins are a sharply-growing emerging technology, projected to grow from a $6.9B market to $73.5B by 2027 at a massive compound annual growth rate of 60.6%. The primary reasons for this steep growth stem from manufacturing: This technology reduces cost and improves efficiency without significant capital investment. Additionally, as other industries like pharmaceuticals and healthcare adopt digital twins, the added demand will pull innovation in the sector and buoy costs through scale economy.
The objective of digital twins is to integrate a virtual model with its real-world physical asset to rapidly assess and improve the asset's characteristic(s). As a result, the manufacturing sector provides one of the most popular applications of this approach. A significant portion of digital twin usage focused on active operation to assess and optimize component performance. Engineers have since applied them to product design and development, commissioning of virtual assets, predictive maintenance, inspection, and life cycle analysis.
Companies have seen substantial gains from extending the use of digital twins across their businesses. Now, city planners are joining the movement by applying digital twins to their municipalities. ABI Research projects over 500 cities will employ digital twins by 2025 to enable mobility and sustainability for their citizens. The rise of electric and autonomous vehicles traversing the streets further accelerates this trend. As a result, using digital twins to enhance the infrastructure that interfaces with these vehicles can deliver cost, energy, and safety improvements to 21st-century cities.
Features of Smart Cities
Urbanization is a global macro trend, with about 55% of the population residing in cities and urban areas. With exponential population growth adding over eighty million people each year, the densely-populated China and India are below the global average in urbanized percentage. These figures signal that the rate of urbanization will increase, with the world's urbanized population projected to approach 70% by 2050. With such a high fraction of the global population in cities, there is a direct correlation between enhancing towns and improving the citizens' quality of life.
Many of the existing smart cities employ several standard interfaces to enhance life for those who live there. Among these are:
- Intelligent light
- Internet and video coverage
- Building air quality and emissions reduction
- Park and common area maintenance
- Public safety
- Sustainability and climate monitoring
- Intelligent transportation systems
Technology will, of course, play a pivotal role in creating smart city infrastructure. Sensors, cameras, and other devices will collect enormous amounts of the currency of intelligent technology: Data.
From there, AI and deep learning process the data into automated enhancements to improve urban functions and add new capabilities. In addition, the permeation of 5G will increase the processing speed by order of magnitude (or more) and reduce response lag, allowing a function to react more naturally. However, there are critical challenges to overcome to achieve these enhancements in a smart city.
Challenges to Smart City Conversion
Despite the many benefits of smart cities, there are significant challenges when city planners decide to convert a traditional urban setting into a connected, smart one.
Infrastructure
The first hurdle is to build the infrastructure required for smart cities. This equipment includes adding substantial sensors and cameras to relevant functions and ensuring the incoming data is resilient, uninterrupted, and accurate. Sensor technology collects and organizes the incoming information for the processing equipment. Examples of the data they collect are traffic patterns. Air contaminate concentrations or the frequency of emergency calls in a given city region.
Not surprisingly, it is much easier to install the data collection and processing technology during original construction than to retrofit an existing structure. Examples of the upgrades smart cities would need are power sources, internet connections, and processor connections, in addition to the added cameras and sensors.
Accuracy
Calibration is a substantial add-on challenge once the city planning team funds and builds the infrastructure. Innovative technology is only as good as its accuracy. As a result, if the input data does not have sufficient resolution for the AI to process the appropriate response, this condition could lead to errant system actions and conclusions about the component's state.
Automotive IoT sensors must detect cars that should be moving but are stopped in traffic vs. those parked on the sides of a road to improve traffic flow in congested areas. As a result, arranging the data in a way the processing software understands is critical. Once the data is ready, an iterative loop managed by the controls strategy feeds physical conditions to the model, then sends model data back to iterate to a solution quickly.
Security & Privacy
The one potentially contentious challenge is the security and privacy of all the data the public infrastructure collects. More data generated equates to a higher risk of cyber-attacks, leading to a question as to whether continually adding intelligent features to cities is the ideal end state.
Digital Twins Solve Challenges for Smart Cities
Digital twins offer significant advantages for cities aiming to overcome these hurdles. Several large cities are already implementing this approach or plan to shortly, including New York, Phoenix, and Las Vegas. Urban areas with heavy traffic like these contribute 60% of global greenhouse gas emissions and consume nearly 80% of the world's energy. This application can illustrate the methods and enabling components required to achieve substantial improvements in this application, both in immediate traffic flow improvement and downstream sustainability progress.
Methods
The first step to matching a virtual model with the city is to create a 3D CAD model of the town. Next, city officials need to digitize as much information as possible to convert the data into a readable form to develop an accurate virtual model.
The numerical software converts the city's division into a network of tiny blocks (or similar geometric shapes) called a mesh to conduct the analysis. Each corner or node of the mesh contains a set of equations bounded by adjacent nodes. The exterior node conditions, called boundary conditions, govern the initial force on the mesh that starts the analysis.
Once the analysis concludes, the virtual city components' behavior describes the physical twin's expected response. Engineers can quickly alter a parameter if the answer is unacceptable and rerun the virtual model to confirm. This method saves significant amounts of time and expense for physical testing. For the traffic congestion example, the virtual model can assess intersection re-designing options or evaluate a new traffic light pattern before putting it into practice.
Enabling components
The components that enable digital twins exist in six primary categories. These steps translate the digital information to the physical world and vice versa:
- Virtual asset
- Data analysis and integration
- Simulation
- Controls
- Connect digital to physical via the cloud
- Measure data to converge and improve the model
This process dramatically reduces the cycle time to converge on the optimal solution, saves the expense of test samples, saves the need to invest in capital to conduct the tests, and de-risks implementation of the change to the boundary conditions supplied from the physical asset.
The virtual asset is the 3D CAD model, for example, the initial state of the traffic picture. Data collected from the physical part can be fed to the model to improve convergence accuracy. From there, the model simulates performance behavior using AI, another numerical simulation such as finite element or computational fluid dynamic analysis, or extended reality overlaid on an image or the actual physical asset. This step avoids the need for costly infrastructure builds. Once the prediction has converged, engineers can upload the data to the cloud to apply to the physical asset. The model should predict the physical component's appropriate response by this point.
Design engineers can use existing modeling software packages to create the virtual model. Traditional IoT sensors are sufficient for data collection, though the camera solutions need a fixed position mode to overlay the model on the city.
Finally, blockchain can deliver the digital identity to link virtual and physical assets together, capturing detailed product information in a highly-secure way. The blockchain contains cryptographic features that ensure safe data transfer and represents a natural avenue for smart city digital twins to improve while protecting security.
Conclusion and Main Takeaways
The IoT enables smart cities, leading to the IoE (everything). Integrating innovative technology into cities can improve citizens' quality of life and the global and local environment through smoother traffic flow design and intelligent lighting.
But while the benefits of smart cities are well known and widely desirable, there remain considerable challenges in infrastructure, accuracy, and security and privacy. Digital twins for smart cities can avoid redundant infrastructure expense using simulation to converge performance with the physical asset. They can also improve accuracy by integrating the virtual and physical components to enhance the model's predictive capability. Finally, employing blockchain in digital twins assures the higher amount of data produced and processed is secure.
As more of the global population moves to cities, smart cities—aided by digital twins—will improve the quality of life for their citizens.