The new Leica Pegasus TRK Neo represents a significant step in the evolution of mobile mapping systems to help quench the global thirst for smart reality data.
As impressive as legacy mobile mapping systems are, it is unlikely that such solutions could meet growing global demand in the long term. High precision 3D reality capture data for road and rail corridors is essential for infrastructure and asset management and even more so for the nascent autonomy market. It will not suffice to do one-and-done mobile mapping – smart 3D reality capture must be kept up to date. The new Leica Pegasus TRK Neo is an example of the next level of mobile mapping systems, one that addresses many of the challenges this geospatially charged future poses.
“There’s about 70 million kilometers of roads and rail globally that need to be mapped,” says Alessandro Nuzzo, Product Manager for mobile mapping at Leica Geosystems, part of Hexagon. “If we follow the concept of smart cities, a controlled way of infrastructure management, and autonomous vehicles, we can agree that it is not happening in significant numbers today, but it will happen in the next 10 years. We can develop more efficient sensors and capture more data at a faster rate, although the limitation is the speed of traffic.” A solution might be more sensors, operated by more people. Today’s systems could make a dent in the demand but require skilled hands – and there is a shortage of technical surveyors. “We needed to create a system that is so easy to use, that even internal staff, apprentices, interns, or non-professionals can collect the data.” This, says Nuzzo, was a big part of the motivation to create the Pegasus TRK solution.
Leica Geosystems has long been a leader in 3D laser scanning products: terrestrial, airborne, and mobile. Their high precision, high productivity Pegasus: One was released in 2013, soon followed by the Pegasus: Two in 2014, a backpack version of the Pegasus, and the Ultimate model in 2018. These have been widely adopted worldwide for all manner of AEC applications by departments of transportation, engineering and surveying firms, and service providers. Applications range from engineering design surveys to energy corridor management to city mapping , and more.
Nuzzo says the design was directly influence by customer experience with the earlier models They were faced with a design criteria dilemma: “Can we do it autonomously? Can we do it in an intelligent way? Or can we do it in a simple way?” says Nuzzo. “And by having that guidance, we could avoid a lot of mistakes, but would be quasi putting the complexity back on the customer.” When they mapped out the legacy workflows, they noted that specialized customers would always figure them out. However, they didn’t think they would work for many operators of varied skills. “We went back and said no,” says Nuzzo. “The customer should only do what they really need to do. That is collecting the data in the most effective way and let the machine do as much of the thinking as practical – this was the only way we would move forward.”
Before we drill into specifics elements of the Pegasus TRK, which are all new and improved components, here are the basics:
While it may seem an oversimplification, the precision of a mobile mapping system owes mainly to the quality of the capture devices – the scanners and cameras – whereas the accuracy is dependent on the quality and tight integration of the positioning components. Precision can be considered in simple terms as how well the captured points fit in relative terms to each other. And accuracy could be considered as how well the points are represented in space. Yes, a group of points can define a feature in high fidelity, but how well does it fit into the scan of an entire city and a geodetic reference frame?
The ”TRK” in the name is shorthand for “transportation kinematics”; and this is appropriate as a substantial focus of the innovation in the system has been engineered to provide precise, robust, and reliable positioning. The standard for legacy mobile mapping systems has been at least a high precision GNSS receiver/antenna, and a tightly coupled IMU (Inertial Measurement Unit). This has served the industry well, but as Nuzzo notes, “There seemed to be an acceptance of accuracy in the range of 10-15 centimeters; something that might be considered GIS accuracy. With the Pegasus:Ultimate, we focused on improving accuracy with great components. And even more so with the Pegasus TRK – 15 millimeters accuracy was the threshold we sought, and this was surpassed.” They did this by not only adding great components but by adding some that have not been standard on legacy systems before.
In a distinct departure from legacy systems, Leica Geosystems has added dedicated front and rear scanners to perform SLAM (Simultaneous Localization & Mapping), primarily to aid the tightly coupled GNSS/IMU solution to improve accuracy. Such scanners are being integrated into autonomy and precision robotics systems with great success, and the Pegasus TRK benefits from this same concept.
SLAM addresses the problems encountered in GNSS limited or GNSS denied environments, like urban canyons, under a heavy tree canopy, or in tunnels. When GNSS is interrupted for more than 60 seconds, the IMU starts to drift substantially. Nuzzo says that this SLAM integration can maintain high accuracy in those situations.
“We learned a lot from our experience, and that of our customers, using our Pegasus:Backpack system, that was designed for indoor use. If you have a 360° field of view with the SLAM, you can better calculate the trajectory,” says Nuzzo. “Because a horizontal field of view works best for SLAM, we mounted them that way. If we had mounted them vertically, you would only see a profile for that spot; horizontal gives a wider field of view.” While the front and rear scanners are primarily for positioning, Nuzzo notes that all of the data is captured, and could be processed into point clouds, as could images from the cameras, for instance, to fill in where something was missed by the main scanner(s).
“We’ve built the GNSS receiver and antenna into the main body of the Pegasus TRK,” says Raphael Goudard, Global Mobile Mapping Segment Manager at Leica Geosystems. Legacy mobile mapping systems typically have a GNSS “rover” attached to an auxiliary post on the system or elsewhere on the vehicle. “By putting a Leica AS11 , 555-channel full constellation GNSS unit inside, we streamlined the system. And we had to do a lot of work with the GNSS and other components, to shield and protect them from interference from one another, for instance, the front and rear scanners.”
Goudard says that having a scanner for SLAM in close proximity to the internal GNSS would have been a challenge due to high interference. How they solved this was to split the field of view between the two scanners, and to keep them at a good distance from the GNSS.
“The internal GNSS made so much sense, and we have an option for a second external antenna for heading if you are mounting this on a boat,” says Goudard. “If you add a second antenna, it has a target on it, so the camera can recognize it and integrate its position into the solution automatically.”
If one is seeking to achieve and maintain the highest accuracy, they may choose to add a DMI (Distance Measurement Instrument) to the positioning solution. Precisely knowing the vehicle’s speed along the path of data capture can improve trajectory calculations.
Optional for the Pegasus TRK is either a mechanical DMI, that fits on a wheel of the vehicle, or an optical DMI. Optical DMI are growing in popularity, as some wheel mount devices are prohibited by regulations in different parts of the world. The optical DMI is typically mounted on a rear bumper, at a known distance from the ground (e.g. 30cm). There is also an optional rail DMI.
A pair of fisheye cameras are mounted to a fixed stem above the Pegasus TRK and angled about 45° from the axis of travel. “To get a 360° image from two fisheye cameras, there can be a bit of a stitch line area,” says Goudard. “Angled like this, you do not see that, this is done in a very clever way.”
Detachable “butterfly” side cameras are each a pair of 24MP stereo cameras. These can be mounted horizontally to read small identification tags on utilities from the images, or vertically to see higher up on trees or powerlines. Goudard says that you might want to position the cameras in ways to help in subsequent meshing of the 3D data if you desire that. A fixed forward-facing camera could be used for many applications, including the drive-through videos that some customers seek. The rear-facing camera is fixed in a position that makes it optimal for pavement condition imaging. A unique feature of the butterfly cameras is the ability to auto-calibrate on-the-fly.
Again, all of the images from these cameras could also be processed photogrammetrically, and the data from the scanners into point clouds, albeit at a lower precision than the main scanner head(s).
The main scanner head, or dual heads on the Pegasus TRK 700 model, are powerful units, as noted in the above specifications. Goudard says that ToF (time of flight) scanners were chosen instead of phase shift scanners, to perform well at longer ranges. While roadway corridor areas of interest for many projects may be much narrower, Goudard says that the rotation speed adapts to the range you desire and can yield precise and accurate data up to 490 meters, on par with many high-end tripod-based terrestrial laser scanners.
He notes that another advantage is the ability to get returns off varied surfaces and materials, like road signs, even thin ones like powerlines.
A key concern with data capture in public corridors and what mobile mapping systems can also see on adjoining private property is of course, privacy. For instance, they needed to design the Pegasus TRK to be compliant with the General Data Protection Regulation (GDPR ) of the European Union.
To do this, the Pegasus TRK has an on-the-fly, AI-driven anonymization algorithm that obscures human beings and their faces and motor vehicle license plates before the images are written to storage. This ensures that at no stage of the processing or data handling are there any stored images without blurred faces and license plates.
For all of its advanced features and components, the Pegasus TRK is lightweight compared to many legacy mobile mapping systems. The TRK Neo 500 is 18 kilograms, and the TRK Neo 700 is 23kg – over 10kg lighter than the Pegasus: Two. Mounting the system on a vehicle roof rack has been made easier with a tilting and rotating track mount. You tilt the rack over the side of the vehicle, attach the Pegasus TRK, attach the butterfly side cameras and any optional sensors, then tilt it up and rotate it along the vehicle axis. The Pegasus TRK also comes with the LOC8 theft deterrence location solution. You can track it if stolen or even lock the Pegasus TRK remotely.
The control unit is a box with an LCD screen on top; it is small enough to fit on a standard car seat. It is essentially a multi-core PC and does pre-processing AI tasks while you are collecting. You can operate the Pegasus TRK via the controller LCD screen or connect any tablet via Wi-Fi and operate it via the web UI – no software to install on the tablet. In Leica Pegasus FIELD you do mission planning (pick an area and it will work out optimal routes), project management, control the data acquisition, configure and initiate anonymization, and access remote support live if needed.
Data storage is on either 2x 2TB or 2x 3.8TB removable solid-state drives (SSD). You can hot swap up to three battery packs; they come with small LCD status screens.
The advanced positioning elements of the Pegasus TRK yield high accuracy, helping reduce or eliminate the need for targets (depending on application). RTK corrections from sources like HxGN SmartNet , a base receiver with an NTRIP caster, or other NTRIP sources, are processed in real-time with the IMU, DMI (if used) and SLAM data to yield high accuracy in optimal conditions. But there may be situations where you would want to post-process the positioning data to refine the trajectory, PPK+IMU, etc. This can be done in Leica Cyclone Pegasus OFFICE .
Additionally, in Cyclone Pegasus OFFICE you can manage projects, classify points clouds, do further anonymization if needed, do some feature extraction, and export to various standard industry formats. There is also Leica Cyclone MMS DELIVER for further feature extraction for road and rail applications.
Shortly after the announcement of the Pegasus TRK, RISE3D in the Netherlands took delivery of two systems – and may add two more soon. Why is it that this firm needs multiple high-precision-high-production mobile mapping systems? Simple: They are mapping the entirety of the Netherlands and will do so continuously. Indeed, a true digital twin must be accurate and current. RISE3D is building a dynamic digital twin, not just a 3D model but also with AI-driven intelligence for everyone.
Raymon Somford, CEO and Founder of RISE3D says that theirs is not a traditional content-as-a-service model like you might see with aerial image vendors. They will not be selling the scan data, but instead, as Somford explains: “We sell the solutions. I came from a civil engineering background, working for one of Europe’s largest transportation and traffic engineering firms. As engineers, we want to do the engineering and not to have to deal with one-off mobile mapping and surveying; we want the solutions.” By solutions, Somford means the end products: the pavement information, the roadway paint markings, signs, utilities, structures, trees, and so much more. RISE3D combines 3D meshed models from mobile mapping with many other data sources like utility records. It offers this in a secure portal where anyone can simply choose the AI extracted, mapped, and attributed features they need for their projects. But the spatial foundation for this rich solution resource is mobile mapped data – a tremendous amount of data.
“We scan and measure while walking, sailing, flying, driving and biking,” says Somford. He and his partners understood that the majority of the public space they are mapping could be most effectively captured with high-precision-high-production mobile mapping systems. “We had about six month experience with a Pegasus:Two Ultimate , and we have a Pegasus:Backpack , and BLK2GO ,” says Somford. “We knew that Leica Geosystems is the expert in scanning and mobile mapping, so the choice for our new systems was easy.” RISE3D immediately put the two systems to work, mapping seven days a week, up to 10 hours per day with each, when the weather is good.
The high precision and reliable positioning appealed to RISE3D, and their tests confirm this. They take advantage of all options, like the DMI. “When we have tunnel sections, or road corridors with tall buildings or trees along them and the GNSS signals may have gaps, we add the DMI to maintain the precision.” In open areas, they are able to rely on the combined GNSS, IMU and SLAM solutions.
One key challenge is data management. “We produce as much as 2.4TB per day or 16TB per week,” says Somford. “Some of the pre-processing occurs while we are driving, like the anonymization. But this carries over into the night with that much data captured.” He notes that they have been testing in conjunction with Leica Geosystems engineers to work out solutions for their data challenges.
The goals of RISE3D speak to the utility of and growing need for systems like the Pegasus TRK: “The current complexity of our overcrowded world requires 3D,” says Somford. “A realistic, dynamic, and measurable display of our environment that will give you the experience of standing right in the middle of reality. RISE3D provides an instant and precise insight into the physical layout of the public space. We create our own digital twin: an impressive virtual 3D ‘Digital Twin City’, rendered accurately with full realism.”
Whereas mobile mappers with legacy systems may have dealt with hundreds of kilometers per month or quarter, the mappers of the very near future may be looking at thousands. All the talk of digital cities, or “metaverses”, will not come to anything without a lot of smart reality data capture. Even at highway speeds, millions of kilometers of roads and rail equate to a lot more mobile mapping than can be handled presently with legacy solutions by orders of magnitude.
The good news is that automation, AI, and advanced components we see integrated so well in systems like the Pegasus TRK means that many more people can be effective mobile mappers than ever before. The grand 3D mapping of the world is just beginning – great for society and great opportunities for business.
© Geospatial Media and Communications. All Rights Reserved.
For an ad-free experience, please Login and Subscribe to GW Prime.