What is a digital twin?
A digital twin is a virtual representation of a physical structure, system, or object. To be a true twin of the physical entity, the digital representation must accurately convey what it is like in the real world. That means it must reflect the current state of the physical entity it represents, and possess the same characteristics.
For example, a digital twin of a building must show details like how tall that building is, what area it takes up on Earth’s surface, and what material it is constructed with. If the virtual representation of the building reflects its surface area prior to a recent addition, it is no longer a true digital twin because it does not match the real physical entity.
Digital twins cannot be “close enough.” While there are some data and models that are relatively accurate and up-to-date, they should not be considered true digital twins if they cannot be used or analyzed in place of the physical entity they represent. A general rule of thumb? If the data is an approximation or estimate, it is not a digital twin.
Examples of digital twins
Given this definition of a digital twin, essentially any physical object can be rendered virtually. Many industries are looking to digital twins as a way to scale their operations. Virtual representations of real physical entities allow more people within an organization to access and analyze information that used to be confined to a specific location. This is especially helpful as workforces become increasingly distributed and remote, and also helps increase efficiency by reducing the amount of physical meetings or inspections required.
One example of how organizations are embedding digital twins into their workflows and strategies is how Tesla develops a digital simulation of each car it produces.1 Tesla’s AI receives real-time data from each vehicle and anticipates upcoming maintenance needs, ultimately reducing the amount of servicing Tesla will need to perform on vehicles and improving the user experience. Chevron has applied the same reasoning to its development of digital twins for its oil fields and refineries.2 Even the healthcare industry is turning towards digital twins to simulate patient health and discover new treatment methods.3
But while digital twins of products are certainly innovative, perhaps the most obvious and useful application of digital twins is the virtual representation of the physical environment itself. Conceptually, geospatial digital twins render all natural and manmade physical structures virtually for a variety of applications and analytics. No company has successfully completed a full geospatial digital twin yet, but many have developed data that form the foundation. These elements are currently being applied across industries.
Digital twins in GIS
Despite the world becoming increasingly interconnected, and travel between locations becoming faster and more efficient, having reliable data about physical places is critical for scaling GIS operations. Across industries, geospatial technicians and analysts are tasked with monitoring physical elements of our world, such as road networks, wildlife habitats, and other geographic features. For many, this requires traveling to different locations and conducting site visits or surveys to collect up-to-date information.
To increase efficiency, many GIS workers have turned to data as a substitute for site visits. But if they are not using an accurate and up-to-date representation of reality, they may be making decisions based on approximations, not facts. Sometimes estimates are sufficient, but many use cases require freshness and precision.
Consider a consumer using a mapping application on their mobile device to locate a convenient place to mail a package; the consumer ultimately just wants to know where the closest shipping facility is to their current position, and directionally how to get there. Even if the resulting point on the map is really indicating a parking lot shared by the shipping facility and a few other businesses, the consumer will successfully arrive at their destination and accomplish their goal. An approximation of the location is sufficient.
But now consider a fleet manager for the shipping company responsible for optimizing package pick-up and delivery routes. With geographic coordinates representing the approximate location of the shipping facility, the route they develop may not take important factors into account. For example, there may be a specific loading dock location the fleet is required to use. There may be limitations as to what types of vehicles (if any) can drive on the road outside of the shipping facility, or what time of day they can drive there. Similar details could also be uncaptured for the previous and subsequent stops on the fleet’s route, leading to miscalculations impacting the overall efficiency. In this case, estimates are not good enough.
Top use cases for a digital twin
Our hypothetical example above is just one way a digital twin could be used to increase efficiency and accuracy in GIS workflows. Every industry has its own unique geospatial applications for digital twins. At Ecopia, we tend to see elements of digital twins applied for the following use cases.
Insurance digital twins
The insurance industry relies on accurate, precise, and up-to-date information about properties. In the past, this largely came from site visits and tax rolls, but today more and more insurers are undergoing a digital transformation with data.
Precision is critical to insurance risk assessment. Whether underwriting a new policy or investigating a claim, insurers need to know the exact location of both the property and any potential risks. If they use data that is just “close enough” they could be under- or over-pricing policies, resulting in increased liability for the company and decreased customer satisfaction.
While a complete geospatial digital twin has not been developed yet, the foundational elements of accurate geocoding, building footprints, and property attributes are already being implemented in the insurance industry. By leveraging these elements, insurance companies can accurately assess risk without needing to visit individual locations. The foundation of a digital twin provides the necessary information for underwriting or claims management, such as the details of the structure and how close it is to hazards. This data is also up-to-date, preventing insurers from making decisions based on old information.
Watch this panel discussion about digital twins in insurance, featuring Tokio Marine, FM Global, and Ecopia AI
Digital twin data for property assessment
Accurate and up-to-date property data is also imperative for tax assessment purposes. As municipalities and tax jurisdictions appraise properties to determine tax rates, they must understand where the property is located, how big it is, and what characteristics it has that could impact its assessed value. As with other industries, tax assessment offices are in the midst of a digital transformation. Many are leveraging computer-assisted mass appraisal (CAMA) technology to their processes, which requires high-quality input data to be effective.
Having the foundational elements of a digital twin as inputs to tax assessment enables jurisdictions to operate more efficiently, reducing the number of field visits required for assessors. Instead of going to physically inspect each property, assessors can view and analyze data about each location in their jurisdiction right in their GIS.
Because properties frequently undergo change that impacts their taxation, having an up-to-date digital representation of the property is essential for tax assessment offices. When data is kept current, reflecting the actual state of each property, assessors can easily identify new taxable activity without having to conduct a field visit. For example, if using foundational digital twin data leads to detection of change in a property that has not been permitted, a tax assessment office can flag that property for further investigation to make sure all possible tax revenue is being realized.
Learn more about Ecopia’s 3D Nationwide land cover, including how change detection can be leveraged to efficiently flag changes at-scale
Stormwater mapping with a digital twin
Climate resiliency is top of mind for many organizations and municipalities. A key part of understanding how to best respond to a changing climate is to accurately model the interaction between natural and manmade features in the environment. Digital twin elements representing these features help climate scientists anticipate and plan for future stormwater scenarios as they build resiliency strategies for their communities.
The impact stormwater has on a specific property or area is dependent on many different factors. The first main difference is between impervious and pervious surfaces. If a surface is impervious, it does not absorb water during a storm event and instead leads to increased runoff. Accurate data about impervious and pervious surfaces enables stormwater planning departments to see where flood events are possible given certain thresholds of perviousness and expected water volume.
The second factor to consider when stormwater planning is the types of surfaces within each category. For example, a paved driveway and the roof of a building are both impervious surfaces, but water falling on each will behave differently. The foundation of a digital twin encompasses these differences and empowers stormwater management offices to accurately model these scenarios.
Without a digital representation of these real-world features, stormwater planning departments might rely on outdated or incomplete data to develop flood models and climate resiliency strategies. While these methods are certainly a step in the right direction, they do not ensure models represent the real world, and do not always keep up with the rapidly changing climate.
See how elements of a digital twin helped the City of Detroit uncover an average discrepancy of $5.6M in stormwater utility fees.
Digital twin of transportation networks
Another way municipalities are leveraging elements of digital twins as a source of truth is in representations of transportation networks. Most communities have a GIS office to manage their transportation infrastructure, and possess some level of geospatial data about these networks. But oftentimes this data is outdated, having been collected a few years ago and not reflecting new infrastructure projects. Sometimes municipalities rely on open source data that is incomplete, not capturing all of the transportation elements they need to accomplish their goals.
As they work towards critical initiatives like Vision Zero (the goal of a future with no traffic-related deaths or serious injuries), transportation planning departments need more than just a high-level representation of features. In the real world, there are many different aspects to transportation and pedestrian safety, such as the availability of crosswalks or the presence of turn lanes. Only by mapping out a digital representation of these features (the foundation for a digital twin) can jurisdictions capture all of these elements with the level of detail and currency needed to allocate funds to infrastructure development projects.
Similarly, planning departments are increasingly turning to accurate and fresh data to plan more sustainable transportation infrastructure. While many datasets of varying quality exist depicting road networks, there are other essential transportation methods and routes available in the real world that are only present in elements of a digital twin. For instance, many communities are improving bike lane infrastructure to encourage more sustainable travel. But to do this effectively requires an up-to-date rendering of current infrastructure that many transportation datasets lack.
See how high-precision mapping data is powering digital twins in the transportation industry.
NextGen 911 & digital twins
Emergency response teams lean on geospatial information as they roll out new public safety measures. However, the way they are leveraging geospatial data is changing due to the shift in 911 calling behavior. While geospatial data detailing associations of landlines to buildings used to be the foundation of emergency response, the societal shift to mobile phones has led many teams to implement elements of a digital twin.
Data components of a digital twin are especially valuable when considering how traditional geocodes and property data only show an approximation of where buildings are located. If a building has multiple floors, emergency response teams may not know where to go to respond to a 911 call. Even though the FCC has now mandated telecommunications companies to provide z-axis coordinates with 911 calls, many emergency response teams have no way of applying that information without a digital representation of the world to map it to. With 3D vector data that accurately represents real buildings, emergency responders can see exactly what floor of a building a call is coming from and expedite services.
The foundational layer of a digital twin can also enable next generation 911 innovation on the ground. It does not matter if someone is calling 911 from a landline or a mobile phone if they are located in a property at the end of a winding driveway - traditional geocoding solutions will route response teams to the street centerline or parcel centroid. Accurate geocodes and land cover data will route them to the building itself, and also detail how to get to the building from the street and driveway.
See how Collier County uses the foundation of a digital twin to send emergency response teams to the right location.
Where to get a digital twin
‘Digital twin’ is somewhat of a buzzword today, and many companies claim to build them. In this section, we breakdown some critical foundational components of a digital twin to help you find the right solution for your needs.
Key foundational elements of a digital twin
The true value of a digital twin is in its accurate and up-to-date representation of the real world. As such, there are a few characteristics that data must have in order to serve as a good foundation for a digital twin.
- Precision geocoding: Geocodes are the foundation of geospatial analysis, so if they are incorrect (even by a few feet or meters) the resulting digital twin is not precise. For a digital twin to be effective, make sure it is developed using building-based geocoding. Some “digital twin” providers use parcel- or street-based geocoding as the foundation for their data, which means they are misrepresenting where structures actually are and how other places relate to them.
- Accurate: If a digital twin is built with an inaccurate foundation, it is not a true representation of the world. It’s important to remember that if data is accurate, it can’t be “good enough.” As you source data for a digital twin, make sure it represents reality. In a geospatial context, that means a 3D building is the exact shape, height, and size as it is on Earth, not just an approximation.
- Up-to-date: The world is constantly changing, and real digital twins will reflect that change. The entire concept of a digital twin is based on building a virtual representation of real physical objects, systems, or structures, which cannot be created with stale data. When evaluating elements of a digital twin, check to see that the information is fresh, otherwise you run the risk of making important decisions off of incorrect data.
- Detailed: There are many datasets out there, but only elements of a digital twin provide the same level of detail you could glean from physically visiting a place. If your use case can be accomplished with only fragments of information, you may not need a digital twin. But if you are conducting complex analytics and modeling of the real world, pay close attention to the level of detail of the data you are evaluating and ensure that it provides the context you need.
Ecopia’s digital twin methodology
Ecopia is building a digital twin of the world using AI, providing organizations with high quality geospatial information. Through our global partnership network, we have access to high-resolution imagery around the world which we then use our AI to extract accurate vector features from in record-time. To give you an idea, we digitized every building in the United States in just six months.
Our partnership network also ensures that we always have the most up-to-date imagery powering our development of a global digital twin. The change detection methodology built into our AI allows us to maintain a real representation of the physical world so our clients can focus on their main use cases, instead of editing old data.
Check out this short video about Ecopia’s methodology for building a digital twin of the world through our extensive partner network.
We can even use your own imagery data as an input to our AI-powered feature extraction. Whichever imagery input you choose, we can use it to extract the real world features that matter most to you.
Ready to leverage elements of a digital twin in your mapping analytics? Get in touch with an expert.