Recharging Drones On The Go With A Supercharger

If Techcrunch is to be believed, our skies will soon be filled with delivery robots, ferrying tacos and Chinese food and Amazon purchases from neighborhood-area dispatch stations to your front door. All of this is predicated on the ability of quadcopters to rapidly recharge their batteries, or at the very least swap out batteries automatically.

For their Hackaday Prize entry, [frasanz], [ferminduaso], and [david canas] are building the infrastructure that will make delivery drones possible. It’s a drone supercharger, or a robot that grabs a drone, swaps out the battery, and sends it off to deliver whatever is in its cargo compartment.

This build is a droneport of sorts, designed to have a drone land on it, have a few stepper motors and movable arms spring into action, and replace the battery with a quick-change mechanism. This can be significantly more difficult than it sounds — you need to grab the drone and replace the battery, something that’s easy for human eyes and hands, but much harder for a few sensors and aluminum extrusion.

To change batteries, the team is just letting the drone land somewhere on a platform that’s a few feet square. Arms then move it, pushing the drone to the center, and a second arm then moves in to swap the battery. The team is using an interesting locking cam solution to clamp the battery to the drone. It’s much easier for a machine to connect than the standard XT-60 connector found on race quads.

Is this the project the world needs? Quite possibly so. Drones are going to be awesome once battery life improves. Until then, we’ll have to live with limited flight times and drone superchargers.

45 thoughts on “Recharging Drones On The Go With A Supercharger

      1. This.

        I was at a trade show where the drones to do this were being demo’d. I asked the company if they thought they’d have problems with people just plain stealing the drones (they hadn’t thought of it). Then I asked them if they’d considered the privacy problems for visually guided delivery drones (they hadn’t thought of it).

        I’m not holding my breath for the Drone Apocalypse. Particularly here where 50kt winds aren’t unusual.

          1. Yes, agreed. I think autonomous RC craft (I refuse to call them drones) have a place in the commercial industry, but delivering your goods is not it. They might be better suited in a controlled environment, such as a factory.

          2. I’m hoping for a network of pneumatic tubes underground or with robotic “mole rats” running through the network bringing me my Chinese food. Elon’ Musk has a great idea with the Boring company for cars but I think much smaller tunnels would be very useful for most stuff. Maybe not as small as the pneumatic tubes that banks use at the tellers, but I don’t think they would not have to be much bigger to be useful. Small objects could be delivered by robotic “squirrels” that run along wires between utility poles, only reaching the ground at drop-off and pick-up sites. A hybrid network of “mole rats”, “squirrels”, and “flying drones” could do the job nicely. If a community doesn’t mind drones then they can use them heavily. If not, then they can invest more heavily in the other options. Regardless, fully automated parcel delivery is quickly going to become ubiquitous and I’m looking forward to it. I want to see the day when I break a mill bit, I throw it in the “trash”, it is immediately recognized for what it was and sent off for recycling. Meanwhile a new one is immediately ordered automatically and delivered wicked-fast to my location and maybe even stored away at just the right place in my workspace by my own logistics robot. … gotta dream, right? :-)

        1. There are already drones that can pickup, transport, and drop payloads. The hardware doesn’t care if its a burrito or a secondary battery with a GPS marker for extended-range swap or recovery, and thus a special mechanism for loading power packs is irrelevant in field situations.

          Hypothetically, theft can be easily deterred by removing the fuse from a crowbar circuit that bypasses the regulators, and LiPol internal protection circuits. Adding random delays would also help ensure the device is placed in a car or home prior to self ignition.

          People have thought of crack-head theft issues, and the fire hazard these devices already offer. Yet the legal issues are something companies tend to ignore unless the settlement cost exceeds the projected profits.

          1. your hypothetical solution solves nothing if i just want the motors, chop them off and leave the rest sitting there. In fact it works better for me in that case because i could trigger it and drive away, allowing the evidence to be destroyed on its own.

    1. For a long time I have been thinking about drones being assigned fixed locations in the sky, like over sports events or crime scenes, or just as advertisements. They would just hover in the sky while service drone(s) would shuttle fresh/tired batteries back and forth between the nearby charge station and the stationary drone. Now I’d like to see someone do that. :-)

      1. That’s basically what a Good Year blimp would do: stay in place and film what’s happening below. Why go heavier than air if a blimp does the job just fine, for a fraction of the energy cost?

        1. So, which would be cheaper to make and would more likely to stay in place in 30mph winds? I agree that passive boyancy is more elegant than active continuous thrust but the downside of expensive Helium, gigantic construction size, and vulnerability to the wind is huge compared to the problems of wasting a relatively small amount of energy. Overall, I would not be surprised if the entire energy costs of using a blimp exceded a continuous operating drone system. There’s no reason why the drone charge station can’t be solar powered (in the daytime anyway) if you care about zero energy impact.
          Also in comparison, which (a drone or blimp) can be deployed in only a few minutes at the sight of interest with no prior opportunnity to plan for it? I really do think drones have the upper hand here. One could even change the design I proposed a bit and fly the unit tethered, running power up the line. (you’d want to light that line up a bit if it needs to be long enough so aircraft pilots can see it) Personally I prefer the battery exchange concept since such a long wire would also be a lightening rod. :-)

  1. Why not use a click-on/off system and design the drone so that a Battery Can be attached either on top and bottom, lower drone, click in battery , Flip drone drop dead battery. Or am I missing Something crucial here ?

  2. Why not use a click-on/off system and design the drone so that a Battery Can be attached either on top and bottom, lower drone, click in battery , Flip drone drop dead battery. Or am I missing Something crucial here ?

  3. “If Techcrunch is to be believed, our skies will soon be filled with delivery robots, ferrying tacos and Chinese food and Amazon purchases from neighborhood-area dispatch stations to your front door. ”

    I imagine Flight of the Valkyries playing.

  4. Putting the battery on the bottom is going to give you a more stable ‘copter since it’s going to be flying with no payload at least half the time.
    Instead of having the battery carousel be horizontal with a lift mechanism, a Ferris wheel arrangement could allow slightly higher pack density. Alternating empty bays and full bays, extract in the used battery, pivot, insert the fresh one, fly away.
    Apart from mission creep why not use a connector more like those on battery powered tools? Lock the battery in place with cams (that conveniently disengage when the ‘copter is docked) rather than the annoying snap clips on many tools.

  5. I was thinking the same thing. Since you already have a smart flight controller on the drone, how hard would it be for the drone to drop the dead pack. Seems easy enough to grab the landing gear and click a new pack into place.

  6. Two ideas:

    1) have 2 half-sized battery packs on each drone; if it is unloaded (just returned from a delivery, so not carrying a package), it could operate with reduced performance on one of the two. This would allow drone movement while one pack is removed and being replaced, which makes it possible to fly the drone to a pack that is ready, instead of a complicated mechanism to handle the packs from a magazine. By swapping out both packs for new, fully charged ones, the terminal voltages should be nearly identical, allowing parallel connection without needing any switchmode converters for voltage matching.

    2) use small but powerful permanent magnets to attach the battery pack to the drone, where the magnetic path is closed to ensure maximum holding force for minimum weight. The electrical connection could either be spring-loaded contacts, or the magnets themselves. To remove the pack, the docking station would use a large electromagnet, positioned and polarized such that is opposes and overpowers the permanent magnets, which will cause the drone to repel the battery pack. Obviously, the magnets would need to have a high coercivity to prevent damage by the strong externally applied magnetic field.

  7. Why not switch out the drone with another fully charged drone? Drone 1 drops off the package on the platform, drone 2 flys off a nearby charging platform and picks up the package, and finally drone 1 takes drone 2’s spot on the charging platform.

      1. The cost would be the same or less if you consider wear and maintenance. You will need to repair and replace your delivery drone twice as often vs having two drones do the route. You are just buying the replacement drone upfront and doubling the service life of both.

  8. I think that everyone in the “tech” world has absolutely lost the plot. People are going in search of problems to their solutions and only looking for the next big thing because they want their 15 minutes of fame and instant riches.

    This whole drone delivery is a key example, its like someone thought “ooh, we have drones, what can we use them for?”. There was no FMEA (https://en.wikipedia.org/wiki/Failure_mode_and_effects_analysis) study no thought of how this might fail at all. This has proven to be pretty common amongst todays advances with some exceptions.

    I wish we as a society could go back to finding a problem and then devising a solution, and the delivering of goods is a problem that already has multiple efficient solutions already in play. In my opinion, the only use that a drone would actually have in delivery is from an automated road going vehicle to the door delivering food, but then again that is based that people are too lazy to just walk out of their houses to the side of the road. As for packages, just have them delivered to your work like everyone else and if you are embarrassed about your purchases then just ask for discrete packaging.

    1. Doesn’t address instant gratification so the next “solution without a problem”, “so over-engineered you need a fifth grader”, will be teleporter development.

  9. Microwave in a satellite dish, add object tracking and a pan and tilt mechagnism. A tuned antenna on the drone, plus a decent rectifier and you can recharge it without having to land… Although government helicopters may not be too impressed.

Leave a Reply to snaslundCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.