Data Accuracy for Digital Twins: Three Real-World Scenarios

In our previous post, we discussed the importance of choosing the right tool for capturing the data of your road network. To sum up briefly, the quality of the base data you capture must be sufficient for the purpose of the data collection. Hence, the hardware that you use must work to a certain level of quality. In this post, we are going to discuss different use cases by which you can decide what is the best solution for you.  

‍

Would smartphone-based solutions suffice for 3D tours in buildings?

In real estate, creating digital twins for 3D tours for clients is more than common. Instead of taking a trip from house to house, prospective homeowners and real estate agents can visit virtually, saving them time and money. The data for 3D models can be captured with smartphone-based solutions that have already proved to be of value for creating digital twins of indoor spaces.  

In the future, these 3D models can be integrated with other information systems to, for example, predict maintenance needs, optimize lighting costs, and develop cybersecurity. The integration probably needs more accurate spatial data than smartphone-based solution can offer today but for 3D virtual tools, the data collected this way is sufficient.  

‍

What is needed for building a bridge?

As with any other construction, in building a bridge the design has to be modeled first. Moreover, the construction process needs to be monitored and assessed continuously to reach an outcome that is structurally safe and pleasing to the eye. The exact requirements for data accuracy can vary between countries but in general, data collected for construction purposes must reach up to millimeter accuracy (like, for example, in Estonia).

The need for such a meticulous approach comes from the purpose that bridges fulfill. Bridges are built to shorten the route and help to get safely over different obstacles, such as rivers, lakes, and highways with heavy traffic. The components of bridges need to fit together seamlessly for the bridge to be durable, beautiful, and resilient against various external factors, such as heavy loads from traffic, environmental impact, and extreme events. Potential flaws in the design model and therefore in the construction process can make bridge elements weaker and result in serious consequences and even, in the worst-case scenario, a bridge collapse.  

The constant monitoring of the building process helps to ensure that the final product is as good as the design. Deviances from the planned works are detected in a timely manner so that delays and mishaps in the construction are avoided. According to the European Union general building rules, the allowed tolerance between the perfect design and built solution is 10 mm. The data used for monitoring and assessment must be accurate so that the conclusions drawn from the analyses correspond to the reality.  

The best kind of equipment must be used for collecting the data needed for building a bridge and monitoring the construction process. For example, equipment that has 1’’ angular accuracy, 1 mm EDM accuracy, 8.1-megapixel resolution, and scanning speed of 26,000 points per second. However, such accuracy comes with a hefty price tag. Nevertheless, the price for saving human lives can never be too high.

‍

How about road maintenance?

Building digital twins for the purpose of road network maintenance calls for high-quality data in terms of geospatial accuracy and image quality. While building bridges requires the highest precision possible, assessing the maintenance needs of your road network through detecting road defects, such as alligator cracking, potholes, depression, and raveling, can manage with less accurate data than that. Instead of millimeters, we can speak of centimeters and in some cases even up to a meter. However, the captured data has to have enough quality for AI-based processing to get the best outcome from the data – fast, accurate, and cost-effective results that have uniform quality.

The data accuracy needed for AI-processing can be obtained by using equipment that uses, for example, LiDAR scanner (+/- 3 cm accuracy), GNSS/INS system, and 360° panoramic camera. With these types of sensors, it’s possible to capture data that reaches a quality level necessary to pinpoint the maintenance needs of your road network, and also save money as they belong to the mid-range expense class. Page BreakIn some countries, such as Germany and Estonia, precise data, in some cases up to the millimeter, is required by law, but in practical terms, a road crew does not need to know the depth or width of a crack in the road down to a millimeter. The evaluation is done on-site by experienced road construction experts. The location of the minor defect is the most pertinent information.

With budget-friendly but less accurate solutions you can make a 3D model of your object and use it for visualizations. With expensive high-tech equipment, it is possible to obtain measurements with millimeter precision to model and monitor the construction of highly complex objects, such as bridges. The mid-range solutions, in terms of cost, can provide you with data that has centimeter accuracy – perfectly suitable for automatically detecting defects and signs of deterioration on road networks.  

We hope to have given you an insight into which equipment will suffice for which level of data quality, so you can decide what to choose according to the purpose of the data collection and the availability of your resources.