Silent Falcon Soars To Greater Heights As
AWS Helps Boost Productivity in Lesser Time
Aerial Imagery Processing Now Takes One Day To Complete Instead Of One WeekWindows Modernization Microsoft
Silent Falcon captures aerial imagery of airports in order to assess whether the runways require repair. They were having trouble with their process, and wasting time uploading massive amounts of data. It was slowing them down, and keeping them from taking on as many clients as they wanted to. After working with Cloud303, not only is their process much faster, but much of their process is automated, speeding everything up and allowing them to take on a lot more clients.
Silent Falcon is a company that captures imagery of airport runways using drones and then processes those images using an application called Metashape. Metashape stitches together anywhere between 12,000 - 30,000 images of each airport they service, depending on the size. It takes the images that have been captured and stitches together what is known as an orthomosaic (defined by ArcGIS as “a photogrammetrically orthorectified image product mosaicked from an image collection”). Metashape then takes this orthomosaic and measures height differences on the runways’ surface in order to detect potholes and other damage. Airports are then able to use the data to secure funding for runway repairs.
Creating orthomosaics is an incredibly compute intensive process. So much data is captured by the drones to be processed by Metashape that Silent Falcon was running into problems with their process. While their drone pilots were each using a fully outfitted, high performance laptop to run Metashape locally, they were finding they needed to keep their laptops plugged in for five to seven days at a time in order to process all the data. This was slowing things down drastically, as drone pilots wouldn’t even be able to start the upload process until after flying home. If errors were found, pilots would have to fly back to the airport they were servicing in order to retake the images. Not only was this slowing down their turnaround time drastically, but this severely limited how many clients Silent Falcon could take on at one time. It just wasn’t the process they had envisioned.
Why Silent Falcon Chose AWS?
Firstly, Silent Falcon was already using S3 for their storage needs and they were having a good experience there. On top of that, AWS had many other options that would work well for them. For example, Cloud303 was able to automate their orthomosaic creation process and scale their compute platform, allowing Silent Falcon to interact with multiple compute nodes while only actually working with the software on one. This made their entire process much simpler, and also allowed them to cut down on their costs.
Why Silent Falcon Chose Cloud303?
Rather than helping them develop an entirely new way of doing things, Cloud303 was willing to take the time to learn how Silent Falcon’s process was already working, and expand upon that to make it better. This was the main appeal of Cloud303 when Silent Falcon was looking for an AWS partner to work with..
Phil Supinski Sujaiy Shivakumar
CEO/Solutions Architect CTO/Solutions Architect
EC2 Directory Service VPC
Cloud303 built a Windows Active Directory based network in AWS in a new VPC that they deployed next to the data they were already storing in AWS. The VPC contains a domain controller (to handle authentication), a file server (for exporting data from S3 to a higher speed solution that the nodes can run off of), a client-server (for building projects), and a coordination server (for coordinating all the projects and sending them to the nodes for rendering) and six rendering nodes, all based on Windows. Cloud303 also streamlined the process by automating the coordination server and rendering nodes, so from the point of view of the user, they’re just using Metashape on a single machine, but in the background, this fast, reliable, efficient system is available. With Metashape ostensibly running as a service, complicated management is abstracted away. Additionally, the system was designed so the nodes could be turned on and off from the AWS CLI using a powershell application written by Cloud303. That way, the fleet would only be running when it’s in use, which helps optimize costs. things, Cloud303 was willing to take the time to learn how Silent Falcon’s process was already working, and expand upon that to make it better. This was the main appeal of Cloud303 when Silent Falcon was looking for an AWS partner to work with..
Now that Silent Falcon has moved their system over to AWS, they are able to process a job in less than 24 hours. They have been able to take on many more clients at one time, and their business has begun to grow in the way that they had envisioned.
Partner Opportunity Acceleration FundingWell-Architected Program