How does mobile mapping work? And how do we do it?

Mobile mapping creates a detailed digital model of reality through an efficient process – from careful planning and data collection in the field to advanced post-processing and analysis. With modern equipment and expert knowledge, the entire workflow can be carried out quickly and with high precision, providing a useful digital twin for planning, documentation and decision-making. To implement mobile mapping in practice, a combination of planning, field data collection and post-processing of data is needed.  

Performing mobile mapping is divided into several steps, this is how we do it:

1. Planning: Before fieldwork, the route and layout are planned. The roads or areas to be scanned are identified and factors such as permitted road sections, traffic, and GPS coverage are taken into account. The route can be scheduled in advance and saved in, for example, KML format to be followed in the field. It is optimal to choose a time with good lighting conditions (preferably daylight with cloudy weather for even lighting) and minimal traffic disruption. Any reference points (e.g. ground support in the form of measured reference points on the ground) can be established if extra accuracy is needed during georeferencing.
2. Data collection in the field: The equipment is mounted on the vehicle (in our case, a rail-mounted wagon for railways is often used). Before starting, the system is initialized – GNSS takes its position, the IMU is calibrated and LiDAR/cameras start collecting data. The operator drives the vehicle along the planned route. While driving, LiDAR scanners send out laser pulses continuously and capture millions of distance points per second. At the same time, the cameras regularly take images (e.g. every 5-10 meters or according to time intervals), which are synchronized with the scan data. GNSS and IMU log the vehicle’s precise trajectory and orientation over time. All data – points, images and position information – are stored on the system’s computer in the vehicle. The collection can be done at a relatively high speed; modern systems can measure at normal traffic speeds (e.g. 80 km/h) and still maintain high accuracy in the collected model. This means that it is usually not necessary to close roads or tracks during the measurement, which minimizes disruption and increases safety. 
3. Post-processing of data: Once the fieldwork is complete, the raw data set is transferred to a processing program. First, trajectory calculation is performed – GNSS data is combined with IMU data to calculate the vehicle’s exact path (x, y, z and orientation) throughout the survey. Often, dual GPS antennas or a local base station or network RTK correction data are used to improve the accuracy of the trajectory. The result is a time-stamped position file for the vehicle. The point cloud is then georeferenced: each laser point (with measured distance and angle) is given coordinates by combining the point’s relative position (from LiDAR) with the vehicle’s position/orientation at the time of the scan. Similarly, photos are placed as panoramic images along the route with the correct geographical position. The point cloud can be colored by projecting the photos onto the points, resulting in a colored point cloud where each point has RGB values ​​from the camera images. 
4. Quality assurance and adjustment: In post-processing, the data quality is checked. There is often overlap between driving distances (e.g. when driving the same road segment from the opposite direction or in several lanes). Strip adjustment can then be performed, which means that different scan distances are compared and adjusted against each other to correct small deviations. If ground supports have previously been placed or there are known reference points in the area, the point cloud can be adjusted against these to eliminate global errors and achieve the highest accuracy. Through these steps, very high precision can be achieved – often at the centimeter level globally, and with millimeter accuracy in detail between points in the same scan. 
5. Delivery and analysis: The final point cloud model and images constitute a digital twin of the environment, which can be used in many ways. Depending on the customer's needs, different end products can be produced. For example, roadsides, signs, pipes and other objects can be extracted from the point cloud to create maps or CAD drawings. Software is available to automatically classify the point cloud (e.g. distinguish between soil, vegetation, buildings, road surface) and to extract features semi-automatically. Common deliverables are: point clouds in standard formats (LAS/LAZ), 3D models/mesh of terrain or facilities, orthophotos generated from the images, as well as GIS layers with measured objects (e.g. road signs, poles) and traditional drawings/cross sections. Data is often also offered via web portals or 3D viewing tools, where users can explore the environment virtually. For example, via a web-based viewer, you can pan in 360° images and click to measure directly in the digital model. Through integration with GIS and BIM systems, the information can then be used for further analysis, design or documentation in ongoing projects. 

The complete workflow – from field measurement to finished digital model – can be completed very quickly compared to traditional methods. The raw data registration (step 3) can often be automated and completed shortly after the field work (within minutes or hours). However, the entire process requires expertise in geodetic data processing to ensure quality. At atritec, we specialize in machine measurement and geodetic analysis, which takes care of the entire chain from data collection to delivery, freeing up time for the customer to focus on the use of the results in their project.

Would you like to know more about how Atritec can help you with mobile mapping solutions? Read more about our services or contact our team for a personal consultation. 

At Atritec, we combine technology and expertise to create accurate digital models – tailored to your needs.