GeoCue's True View Ecosystem
Author: Lewis Graham
Published On: October 15, 2020
Analyze the Drone LIDAR Market
The drone LIDAR space is becoming a bit crowded with a number of systems on the market ranging from truly terrible to remarkably good. Of course, GeoCue Group strive to be in the remarkably good category. The core requirement for this is to produce an easy to use, reliable sensor that meets your project requirements. Fundamental to these requirements are, of course, the performance characteristics of the system. Does it meet the accuracy, precision, density and so forth envelope needed to generate the products requested by your stakeholders? Equally important is the consideration of the overall workflow necessary to profitably move through the process from acquisition to finished products. In other words, you need a total solution, not just a sensor. If you evaluate the options available in the market today, 90% of the current contenders are eliminated because they leave you to your own solution once the sensor has landed.
The True View Ecosystem
When we were laying out our True View plans in 2018, we did not just conceive the True View 410 sensor; we worked on mapping out an entire True View Ecosystem. In my use of the word ecosystem, I mean "economic" system; you have to be profitable (even if you are a government or not for profit organization). We have many years of experience with both building hardware systems (part of my team was integral to the design of the Z/I Imaging Digital Mapping Camera) and complete workflow software. After all, we are the company that brought the GeoCue Earned Value Management (EVM) data processing system to market!
The True View Ecosystem is implemented as three distinct components (see Figure 1). These components are:
- The sensor itself, of course, along with its on-board software ("firmware")
- True View EVO, our post-processing and analytics desktop application
- True View Reckon, the Amazon Web Services "cloud" component of the True View Ecosystem
Figure 1 : True View Ecosystem Diagram
One of the major design decisions we faced is the definition of the so-called "edge." This is the boundary between functions locally performed and those relegated to the cloud. We closely monitor the overall drone mapping industry, taking note of what seems to work well and designs that have proved disastrous. For our industry, requiring a direct link from the sensor to the internet has just not worked. Obviously the designers of these schemes have never been to a real mine site! For this reason, we elected to have a hardware transfer of data from the sensor to a desktop processing system and involved the cloud solution where it truly makes sense.
I will give you an example of this. We offer our True View 410 3D Imaging Sensor (3DIS®) under a subscription model. We teamed with Applanix (a Trimble company) to "meter" the Position and Orientation System (POS), allowing us to amortize this significant hardware cost over flight minutes rather than a total system upfront payment. One of the design decisions was how to authenticate the sensor. One approach is to have the sensor contact a cloud service on bootup. This would require a quasi-real time connection with some sort of authentication service. The typical way we have seen this implemented is to have the sensor communicate with a local device such as a tablet (usually via Bluetooth) and that tablet, in turn, communicates with an authentication service over, you guessed it, the internet. No connection means no flight. Sticking with our firm True View Ecosystem rule of no external communication requirements while on site, we moved authentication to True View EVO, the desktop publishing software used for post-processing.
Cloud Hosted Services
Cloud-hosted services are really attractive to vendors (including us) because they offer an opportunity for annuity revenue - they provide reliable streams of revenue to a company without the need for constantly selling a new, discreate local product. I like this model as well as anyone but it has to make complete sense to our customers. If cloud processing offers me an advantage, it has to offer you an advantage at least as good as mine!
In keeping with this value equation approach, another rule for the True View Ecosystem is that paid cloud services cannot be a requirement for using the system. This means that core functions within Reckon are not metered. Examples include calibration file management and sensor health monitoring (more on this later).
To be successful with small Unmanned Aerial Systems (sUAS) data collection and processing, you really need post-processing software that is co-engineered with the sensor. If you are shopping for an sUAS LIDAR system and the purveyor suggests a third party product for post-processing, think twice before you proceed. We developed True View EVO, our post-processing software, concurrently with the sensor development. For example, our True View 410/615/620 systems are not just LIDARs; they have integrated photogrammetric cameras that were carefully designed to provide rigorous fused data. We call this class of system a 3D Imaging Sensor (3DIS). But what good is it to pay all this attention to careful hardware design if there is not an equally capable post-processing software that can extract maximum value from the data? Our True View EVO software has a number of features that take advantage of the sensor configurations. Perhaps the best example is our fusion of RGB imagery into the point cloud. Most systems colorize the LIDAR point cloud from an orthorectified image created from the photos, independently from the LIDAR system. This has several disadvantages. The first is that it requires you to first make an orthophoto and then do a colorization process. This adds significant time to the post-processing workflow. The second problem with this approach is that considerable data value is lost. Colorizing a 3D point cloud from a 2D orthoimage means that vertically stacked features (imagine ground with an overhead wire shaded by tree canopy) all get a single color value (usually the color of the top object). True View EVO does direct tracing of 3D points to the original source image. Not only does this provide true 3D colorization but it also makes the workflow considerably faster.
We use True View Reckon for functions that directly add value. For example, all True View systems use Reckon to update the sensor calibration file. In analyzing where things go wrong in processing sensor data, one of the prime factors was using an incorrect calibration file. When post-processing in True View EVO, the serial number of the sensor is transmitted to Reckon. Reckon then looks up the appropriate calibration file from its repository of files and transmits the calibration file back to True View EVO. This system is calibration epoch aware, providing the correct calibration file for legacy projects that might be using an older calibration.
Soon to be released (by November) is a new sensor health check monitoring service. True View 3DIS systems record an extensive log file, the Cycle Log, during flight. This file contains environmental, processing and health information about the flight and the sensor. The Cycle Log is one of the Layers formed when you import a flight into True View EVO. A graphical display is provided with geospatially correct one second "heartbeat" points that can be clicked to read log data relevant to that time slice - very cool and useful! However, subtle things can happen in a sensor that do not damage the output but may be indicators of an issue. We will soon be pushing the Cycle Log files up to Reckon as a background operation while you are post-processing in True View EVO. Algorithms in Reckon will analyze the log and, if something is amiss, automatically send an email to our support staff, along with the log file. I am very excited about this new proactive monitoring. Reliability is critically important - these are smart ways to approach improvements. This service will be included in both subscription and purchased systems at no additional cost.
Finally, of course, are paid services offered via Reckon. A recent addition to True View EVO is the ability to store and retrieve Cycle files to and from True View Reckon. This provides you with an easy backup solution for safely archiving project sensor source data. The fee for this is economical and charged, like everything in Reckon, via True View Points. Beyond archiving, it is also useful for transmitting data from field crews to your back office processing center. Field crews can kick off posting a Cycle to Reckon at the end of a flight day and your processing technicians can download it in the morning. One big advantage of this approach is that all files are nicely bundled into a single compressed archive.
Present Your Product
Another very useful aspect of Reckon is presentation of products for end users. An end user can log in to their Reckon web site and visualize a mapping site (MSite). - see Figure 2. This is especially convenient for repetitive projects such as monthly volumetrics at a mine site. Reckon is temporally aware, allowing a user to view a site on a specified collection date. Reckon also provides a few tools aimed at specific project types such as the ability to generate a pdf volume report or download a CSV of volumes for the accounting department.
Figure 2: A Mapping Site "MSite" depicted in True View Reckon
We are constantly adding value to the overall True View Ecosystem. We have a number of new capabilities we will be rolling out over the next few software cycles that I think you will find quite useful. I really believe that our holistic approach to sensor hardware and software integration provides true value to you, our customer.