Tag Archives: accuracy

Introduction to Automated Mobile Survey

Dull, Dirty, and Dangerous. Known as the 3Ds of robotics, they describe conditions unfavorable for human work, but ideal for robots. Coincidentally, dull, dirty and dangerous characterize many of the environments where another “3D” (three-dimensional terrestrial laser scanning) is used.  It should come as no surprise then to see robotics and 3D laser scanning united in purpose… that being, large scale 3D survey.

The team at Allpoint created Automated Mobile Survey (AMS) to maximize the performance of existing laser scanning tools. Our aim is for AMS to boost scanning productivity by allowing users to do more work with less resource.  Most users of static scanners already have well establish practices and workflows for collecting scan data and ensuring accuracy of registered point clouds.  Those workflows usually entail

  1. installation of targets;
  2. moving of a scanner from station to station;
  3. registration of scans; and
  4. removal of targets.

These workflows are tedious, time consuming and add little value to the overall delivered product.   AMS was conceived to alleviate many of these burdens.  Specifically, our goals are to

  • make 3D data acquisition easy,
  • generate results of survey quality,
  • provide the ability to tie into control,
  • fit within existing workflows and software tools, and
  • increase the productivity of scan collection.
Automated Mobile Platform

Automated Mobile Platform

To achieve these goals, we automated as much of the acquisition and registration as possible. We first sought to reduce the amount of manual interaction required to move a scanner through a space. Our Automated Mobile Platform (AMP) is a self-guided mobile platform that carries and operates a 3D scanner and is programmed to follow a specific path for the purpose of collecting scans automatically. AMP is first walked through a scan area to create a “base map” of the space.  The base map is a 2D floor plan allowing operators to specify a drive path and scan locations.  When complete, AMP is ready to start work.

The AMP, when used in conjunction with Allpoint’s cloud-to-cloud registration tool, Scan-time, constitutes the entire system behind AMS. As the AMP takes scans, data files are stored and relayed to an off-board laptop running Scan-time. Coordination between the AMP and Scan-time allows data collection and registration to occur simultaneously and automatically in the field.

Automated Mobile Survey

Automated Mobile Survey

What about areas where wheeled systems cannot go? The AMS is a fully integrated solution. Static tripod-based scans are easily integrated into the registered scan set.  Static scans collected in the same operating area as the AMP can be imported into a common repository and registered live in the field.

As part of a fully integrated solution, AMS has the capability to tie into a control survey. Using Scan-time and a few survey targets, registered scans are linked to survey control. Scan-time provides accuracy reports detailing the quality of registration and control.

Finally, AMS provides features that enable quick review of scans collected in the field via Insight, a hosted, web-based portal accessible from any modern web browser or tablet (including the iPad). As soon as scans are collected, they can be uploaded to a server allowing project managers or stakeholders to review, measure, annotate and share scan imagery. 

Insight

Allpoint's Insight

In conclusion, we here at Allpoint are pleased to offer the laser scanning community with a new set of tools that will enable you to be more productive, with less resource, and no sacrifice in quality. We are technologists with backgrounds as services providers and a passion for laser scanning. We have built our tools to help others contend with the dull, dirty and dangerous scenarios of scanning in harsh conditions.

FAQ

Q: Do I need targets?

A: Scan-time requires no targets for registration; however, a few targets (minimum of 4 over the entire job) are necessary to tie the registered scan collection into survey control.

Q: How do I tie AMS into control?

A: Specific details are outlined in the Scan-time documentation, but the overall process is straight forward. A minimal set of targets are placed into the area where laser scans are collected.  These targets are surveyed, extracted, labeled and placed into a simple table.  Scan-time then imports this table, aligns the scans to control and provides an accuracy report. 

Q: Are results from AMS survey quality?

A: In the environments where AMS is currently being used, which include manufacturing facilities, industrial plants, warehouses, office buildings, steam tunnels, museums and airports, the results have been evaluated and shown to exhibit an RMS error of 3mm to 5mm. 

Q: Are the mobile platform (AMP) and registration software separable?

A: Yes.  AMP is only the automatic collection platform. In some jobs, AMP as well as any other wheeled system cannot access congested spots or is not appropriate for the work. In those situations, Scan-time can still provide significant productivity gains and in-field registration through its 4-click cloud-to-cloud registration interface.

Q: What scanners are supported?

A: We build our products to allow you to mix and match scanners. Currently we support native file formats for Z+F, FARO, and Leica products with Surphaser and Riegl in the very near future.

Q: Can I register multiple scanners together at the same time?

A: Yes.  We frequently combine scans collected from the AMP with scans from a tripod. AMP is ideal for covering long hallways or open areas, but tight spaces such as closets are difficult to reach.

Automated Mobile Survey™, AMS™, Automated Mobile Platform™, AMP™, Scan-time™, and Insight™ are trademarks of Allpoint Systems, LLC.

iPad®  is a registered trademark of Apple.

Mobile LiDAR’s ‘Tipping Point’

As Joel has mentioned previously, we recently returned from a trip to the 90th Annual TRB Conference in Washington, DC. For me, the goal of the trip was to assess the state of the Mobile LiDAR industry as a whole and get feedback on some of the concepts we hope to introduce in this space.

First of all, many thanks again to everyone who took the time to meet with us, both in our Focus Group and individually; your expertise and input will help to shape our offerings as we progress in the development cycle.

The Mobile LiDAR industry is at a fascinating place in its maturity. At this point (a few years in), the initial benefits of Mobile LiDAR are beginning to be seen (safety and efficiency at the project level). But, we have yet to achieve the ‘tipping point‘ that will push the technology into the mainstream. The key to achieving this push is recognition of the value of the technology at all levels of an organization. Mobile LiDAR needs to be part of corporate and public best practices, not just a tool in a project-level toolbox.

I came away from the trip with three themes that I feel will be critical in taking the Mobile LiDAR industry to the next level:

  • Tailoring LiDAR data collection to intended use

There was much talk during the conference about the various levels of quality (‘Survey Grade’, ‘GIS Grade’) that can be achieved with Mobile LiDAR. I get the sense that some customers hope to save money by relaxing accuracy requirements. For example, a DOT who needs to perform a sign inventory may only specify a LiDAR accuracy requirements of +/- 2′. This allows the surveyor to cut corners and submit a lower bid for the project.

This approach is problematic, as it can severely limit the value of the collected LiDAR data to the organization. An organization can only achieve maximum value from LiDAR if it can re-use the collected data across multiple projects, teams, or corporate levels. If a future, related project requires a higher level of accuracy, the previously-collected data could be worthless. End users should look at Mobile LiDAR as a data source which can be continually mined for different information.

  • The need for tools for communication and collaboration

It is ironic that LiDAR data in essence models reality (the ‘as-built’ world), but most users still resort to CAD models, texture-maps, and other decimated representations to communicate and collaborate with others about a project. This is rooted in the fact that 3D tools can be hard to understand for an untrained user; further, the huge data sets common in this industry make transferring and visualizing LiDAR files very difficult. There is a clear and present need for an enterprise-scale system for sharing and collaborating with large LiDAR data sets. Check out Joel and Aaron’s talks at the ILMF and StreetMapper conferences for more on this.

  • The mystery of accuracy

It seems that I everyone I talk to in the Mobile LiDAR field has not only a different definition of accuracy, but also a different way of measuring it. Proper assessment of the accuracy of each component in the system (control, GPS, LiDAR measurements) is critical for both proper project specification and correct generation of down-stream information products (e.g. vertical bridge measurements, contour assessments, etc). Some standardization efforts have occurred at places like NIST, but there is a long way to go before we are all on the same page when it comes to accuracy.

As platform-agnostic software providers, this issue is critical to Allpoint because we need to ensure a baseline quality for the data submitted to our system. Further, many extraction and measurement algorithms require knowledge of measurement noise/accuracy to perform optimally. As a Mobile LiDAR user, accuracy should be important to you because you need a standard way to specify requirements to hardware, software, and data providers (surveyors), and to verify the results you receive. You need to ask the right questions and be able to understand the answers. Stay tuned for more on this subject.

Ryan Frenz, Head of Software Development