Cape is an aerial telepresence platform, designed for the ultra-low latency control of UAVs from remote operation centres. What does this mean? Cape allows users to fly a drone, from the other side of the world in real time! Use cases include emergency operations, asset inspection and virtual tourism.
I had the opportunity to fly several CAPE missions at Comms Connect 2018, using RGB and thermal cameras in Ensanada, Mexico.
Telstra Ventures made an investment into Cape in early 2018, and my colleague Peter Jones (UAV Program Lead) has been working on Proof of Concepts that leverage the capabilities of CAPE, with the view to use CAPE both internally, for our field force, and to inestigate how we might create new products and services for Telstra.
The secret sauce of CAPE lies in the low latency, in the order of 100-200 milliseconds, which is is enabled over Telstra’s mobile network.
Night missions are enabled through the use of thermal cameras.
Different thermal colour gradients are also possible.
CAPE has also tested the integration of artificial intelligence algorithms into their platform. The AI is used to augment the video feed being shown back to the pilot, or assist with control of the craft.
Other more future state applications include facial recognition or license plate recognition. This may be enabled by having the camera automatically zoom to each person’s face in the shot, and take a photo for comparison with a database.
More immediate applications include lost person detection (Search and Rescue) and anomaly detection (Asset Inspection).
An AI application from CAPE which recognises vehicles, and autonomously tracks them using the gimbal.
Police in Mexico are actively using CAPE, carrying drones in the boot of their squad cars, allowing them to get aerial surveillance at a moment’s notice. Pilots are able to remotely connect to the craft, allowing the police officers to continue with their duties, and experts are able to remotely view the feed and provide insights.
Similar applications exist for asset inspection- a cell tower technician is able to travel to a remote site, and put a drone in the air. The engineer is then able to view the feed in real time, and point the camera at the specific areas of interest. This mitigates issues for asset inspection where the captured data misses key data, and the flight needs to be carried out again.
Another example of asset inspection would be long flights for pipeline inspection. AI is able to scan the feed for damage, across 100s of kilometres. When damage is detected, the pilot is then able to take the UAV in for a closer look, enabled by the real time control, as opposed to requiring a second flight.
Search and Rescue
For Search and Rescue, CAPE would allow a swarm of drones to be flown simultaneously, with an AI algorithm scanning the video feeds for lost persons. Once a drone makes a detection, the human operator is able to jump into the “cockpit” and validate if it is indeed the missing person.
Lastly, CAPE enables virtual tourism, where users control what they see, and where they go, getting a bird’s eye view of key tourist attractions
CAPE's commercial video demonstration.