An inspection takes place during and after you fly your drone. Our video player lets you review your asset frame by frame and bookmark what’s of interest.
Your inspection data is mapped into a sparse 3D model of your asset so you’ll always know where your points of interest belong.
Points of interest can be documented, classified, measured, adjusted, overlaid with thermal data, and all of this can be easily packed into a report.
“Inspector 3.0 and its data localization feature really helps me get everybody on the same page. You get the data necessary to get the job done!”
– Simon Kumm, InspecDrone GmbH –
“The localization feature of Inspector is really what was missing up until now. It was always asked by my customers. Now I can say, look it’s here.”
– Thibault Chatillon, Innovation and Development Manager at Setec –
The mapping feature of Inspector 3.0 relies on Visual-Inertial Odometry (VIO). VIO estimates the position and velocity of Elios 2 by using only the input of the main camera plus the Inertial Measurement Unit (IMU) attached to it. VIO is a viable alternative to GPS and lidar-based odometry to achieve accurate state estimation.
For the moment we cannot certify the accuracy of the mapping algorithm and data localization feature of Inspector 3.0. Also, accuracy varies according to the environment where the drone is flown — the presence of dust, reflectiveness and amount of visual features of surrounding surfaces — and the care brought by the pilot in following piloting best practices like maintaining good image quality and avoiding big impacts. However, out of the flights that were successfully mapped we observed an accuracy in the range of half a meter when conditions were not great down to 10 centimeters in ideal conditions (20 to 4 inches). In general, this accuracy was sufficient to bring a lot of value to the user in the context of their use case. Again, we give no guarantee that you will observe the same accuracy and it is actually one of the objectives of the beta tester program to better characterize the accuracy of the mapping algorithm. So, we want to hear about your experience.
Photogrammetric 3D models are meant to be extremely accurate and to combine visual features (texture) with 3D point clouds or 3D meshes. With a photogrammetric 3D model you can perform volumetric measurements and add Ground Control Points to improve the quality of your model in order to make it as close as possible to reality (digital twin). All of this is at the cost of acute piloting skills, processing time, the amount of data required, and deep knowledge of the software used to produce the model. The mapping feature of Inspector 3.0 is meant to localize your inspection data by producing an accurate trajectory with a sparse point cloud offering a good balance between accuracy and processing time. It does that out of regular inspection flights and not ones dedicated to optimizing data capture for photogrammetry.
Yes. There is not an export button per se but the outcome of the mapping algorithm is a standard *.PCD file along with *.TRAJ file for the drone trajectory which can be displayed and processed by other software available on the market.
No hardware update is required, however, you’ll need to update the firmware of your drone to the latest version. The Elios 2 is equipped with a camera and an IMU which are the only elements required for the VIO technology to work. Thus, the mapping feature of Inspector 3.0 is readily available to all our Elios 2 clients.
OS: Windows 10 (update 1903+ for Pro and Home editions)
CPU : needs to support virtualization, tested on Intel Core I5-6600, I7-6700, I7-7500
RAM : tested with 8GB+ (less than 8 GB may not work)
Graphic Card : not required
Yes and no. The mapping algorithm will work on the vast majority of the flights you will perform. However, the speed at which you fly and the care you apply in flying smoothly without having the cage (even partially) in the image is directly related to the density and the accuracy of the point cloud. So, depending on the density and accuracy of the model you need you will or will not have to adapt the way you perform your inspections. In any case, here are the 3 golden rules to follow to improve your data capture specifically for Inspector 3.0’s mapping algorithm :
1. Good image — camera within 15cm (6in) to 10m (33ft) from a well lit surface
2. Loop closure — Take off and land at the same spot looking at the same stuff. If flying in tunnels, do regular – every 30-50m (100-160ft) or so – 360º rotations to force loop closures
3. No cage
Unfortunately not. The mapping algorithm of Inspector 3.0 does not successfully work with Elios 1 video stream due to the cage obstruction. Therefore, Elios 1 is not supported by Inspector 3.0. Users of the Elios 1 can rely on Inspector 2.2 for their data post-processing.
Yes, you can open a project which has already been processed and visualize 3D point clouds and locational data with the free version of Inspector 3.0. However, you cannot process flights to create a 3D point cloud and locational data with the free version of Inspector 3.0. So, if you intend to share data with your clients or colleagues they don't need to buy a premium license to visualize this data. The free version will do. Ultimately, anyone who downloads the free version of Inspector can process a dataset captured with a drone for which there is an active premium license.
The free version of Inspector 3.0 lets you visualize and document flights. You can also visualize projects containing 3D point clouds and locational data that have been processed previously with the premium version of Inspector 3.0. However, if you want to test the mapping function on data you have captured with your drone without a premium license, you won't be able to do so. For that, you need a premium license and the 30-day trial is a free way to test the mapping function on your data to test it out before committing to buying a one-year license.