Major delivery firms are experimenting with drones for deliveries. If you've started to experiment with drone flying for fun, you'll know that it isn't easy. There's a steep learning curve, and they crash quite a lot, I've found.
However, it doesn't have to be like that. Collision avoidance is technically feasible. The drone senses objects and diverts.
One of the issues with that, though, has been that developers only have a certain amount of time and money to develop new features. Those features have, thus far, been the tempting, seductive ones, like "follow-me" where the hobbyist drone follows the pilot like a dog on a leash. It's the latest thing.
Addressing the propensity for drones to crash is, quite frankly, a bit expensive, and a bit boring. Marketers haven't bothered too much about it.
No-fly zones are beginning to be installed by drone-makers as a token safety gesture.
However, understanding where the population is at any one time and coupling that data with the needs of the sender and receiver of the package, using algorithms, may be a better idea, according to Antton Peña, a student at London's Royal College of Art.
Peña was exhibiting at SkyTech, a commercial-oriented drone expo in London that I attended recently.
Peña has developed a project called Flock, which he says is inspired by clouds in that it analyzes what's happening above our heads. He says that a pro-social flight regulation policy needs to be developed.
Pro-social Flock would determine where the population is and route flights around people. This would alleviate "fear of accidents and undesired consequences," Peña says. That kind of thing, if allowed to happen, might derail drone delivery altogether.
Blanket no-fly zones don't make sense, Peña thinks. The idea that a country or city forbids any use of drones is absurd when, with gathered intelligence, you could easily figure out safe flight paths.
Peña uses the example of a public park, which will be closed at night and consequently empty. Why should there be a blanket ban on over-park flights, as there is in the eight 5,000 total acres of Royal parks in London, due to privacy and safety reasons, at times when there's no risk?
Peña's algorithms, which he's developing, would take into account the fact that the park is empty at night and route the flight through the area at that time. That same people-avoidance concept could apply to other forms of commercial drone use, like utility-line inspections.
Flight Impact Level
The algorithms work using an impact level that is adjusted in real-time, based on how many people are around. He calls it a Flight Impact Level (FIL).
Laws could regulate flight based on this calculated, weighted number, rather than imposing blanket bans.
Sensors and smartphones collect the data in real time, and the FIL changes frequently. A financial district, for example, is often empty during the weekend, but busy during the week. The key to the whole thing working is getting enough sensors participating in FIL.
Using Flock, senders can then schedule their deliveries based on the FIL algorithm. Flockbrain, another element, determines flight paths with waypoints, which it loads directly into the drone.
Fleet managers control the whole process with an app called Flockguard. Recipients and their respective shipments are manually entered into the app, and the algorithm does the rest, including provisioning foldable, wall-mounted physical perches, for accepting packages, which are located around the city. The perches could double as people sensors.
The elegance of the whole system is that by knowing the level of disruption that a flight might have to address before it takes off, the drone doesn't have to react to obstacles or citizens. They aren't there to begin with.
Route planning creates efficiencies too, and if Peña is correct, collision avoidance systems—the expensive future odds-on technology—could become virtually obsolete.
This article is published as part of the IDG Contributor Network. Want to Join?