One of the challenges for drones flying geospatial missions in remote areas is the lack of high quality Internet connections.
This particular issue was part of Amazon’s research into drone-based parcel deliveries in the UK two years ago. At that time, Amazon’s investigation concluded, “Although they won’t be controlled directly by a human operator, the delivery drones will need to be able to receive air traffic control information and instructions in real-time…. Unless regulators set aside dedicated spectrum, drones will probably have to use a combination of Wi-Fi and cellular networks to maintain Internet connection. This will make them vulnerable to network congestion, as well as eat into battery power. The deployment of new low-power, wide-area networks will certainly help, but a key question for the UK trial is how long a drone can safely be without an internet connection.”
The same is true for drone-collected geospatial data storage, pre-processing, and ultimately, shipment over Internet from points in the field.
To address this persistent issue, Amazon and other companies have continued to work on solutions. The result has been an ingenious idea that combines older IT methodologies of storing data on devices that are locally placed in distant geographies and offer a temporary resting place for processing and preparation of data, and movement of the data in a more compact form that requires less bandwidth to a state-of-the-art Cloud environment.
This is a creative use of “store and forward” technology that gets at the crux of many field-based drone missions.
As an example, today’s high resolution cameras can take photos at a granularity that exceeds 2 million pixels per photographic image. If a high res camera on a drone is photographing a landscape at a rate of one photo every 2 seconds for a duration of 30 minutes, that’s a sizable data payload.
“We recognized that there had to be a practical way to collect all of the geospatial data companies needed in the field, and that the method couldn’t necessarily rely on real-time Internet connectivity for data transport, especially in remote locations,” says Rahul Thakkar, director of Commercial Cloud for Insitu.
Insitu should know. It is a wholly owned subsidiary of Boeing, and specializes in aerospace and unmanned aerial systems (UAS).
Here’s how the local storage-then-to-Cloud concept works. A drone is equipped with an Amazon Snowball Edge (AWS), a 100TB data storage and transfer device with on-board storage and compute capabilities. The device can move large amounts of data into and out of the AWS Cloud, and it functions as a temporary storage device for large local datasets or to support local data workloads in remote or offline locations. The beauty of the hardware is that it only costs around $300 per storage unit, so it is affordable.
Once data is collected on the device and ready to be transported to the Cloud for more permanent storage and for use with analytics, Insitu uses its INEXA Cloud that is hosted on AWS to assist clients with analytics and information management on the Cloud-based data.
“Thanks to this collaboration [with AWS], customers can easily and quickly move huge amounts of data to the Cloud, avoiding the expense of building and maintaining high speed networks or building, securing, encrypting, maintaining, and shipping rugged storage devices and computer systems,” Thakkar says. “This can solve a major problem that customers operating in remote areas have faced on a daily basis.”
Does this solve every area of drone-based data collection in remote areas? Not quite. Companies still have to prepare their data before shipping it to the Cloud over the Internet.
Major steps in data preparation include:
- Deciding which data you really need and which you will discard;
- Getting rid of data duplicates;
- Correcting data inconsistencies and/or filling in areas of data that are missing; and
- Eliminating pieces of data that are irrelevant for the project you’re working on.
The goal is to end up with a data file that is substantially smaller than the unedited data that the drone originally collected. This more compact data can easily be transported as a data payload over the Internet, where it will be operated on with specialized analytics software such as the one provided by Insitu.
“We wanted to develop a Cloud-based solution that could also combine with temporary local storage that would allow companies to pre-check and pre-validate the raw data that drones collect before the data was transported to the Cloud for use in analytics,” Thakkar says.
How much does a hybrid local storage-Cloud storage technology like this help companies that collect and use geospatial information?
There will always be companies whose business demands real-time geospatial data coming from the field. In these cases, companies are willing to spend for the extra bandwidth, networks, etc., that might be needed to support the effort. But for the vast majority of surveyors, mining companies, agribusiness companies, and others using field-based drones, being able to temporarily store and pre-process geospatial data at a relatively low cost – and then being able to ship a subset of this data to the Cloud where it will be easier to run analytics against the slimmed down data – is a cost-effective and relatively straightforward deployment option worth considering.