Copters Post Mortem

[mikeal]

When I started putting together a supply list for the copter session I actually found it quite hard to know what to buy in terms of extra parts and batteries and even knowing the number of copters.

I'm going to post the numbers that I ordered here which we can combine with a list of breakages so that future copter event organizers might get a better idea of what to order.

A few things to know about our copter sessions.

• We rant the drones pretty much consistently for 8 hours.
• Every hour a new group would come in and get flying. This meant that the average time spent flying vs. coding was much higher than previous NodeCopter events where people spent 4-8 hours with a single drone.
• The barn had very high ceilings and some pretty high rafters which contributed to the types of breakage we saw.

Here's the list of supplies I ordered:

14 x Parrow AR Drones 1 x Parrot AR Drone 2.0 Central cross 1 x Parrot AR Drone 2.0 Gears & Shafts Set of 4 1 x Parrot AR Drone 2.0 Motor and Controller 3 x Parrot AR Drone 2.0 Tool Kit 4 x Parrot AR.Drone Replacement Propellers 40 x Parrot AR.Drone 2.0 Battery Lithium-Polymer Replacement Battery 20 x Parrot AR Drone 2.0 Charger Set 5 x Kensington 62690 SmartSockets Standard Adapter 8-Outlet Color-Coded Power Strip and Surge Protector

Those surge protectors are a find because they are the only power strips on amazon that can fit a drone charger in every slot.

In addition, Walmart Labs brought some drones and I believe some supplies, I'll let them comment on what they brought.

[mikeal]

@aulvi and @nexxy and @nvcexploder : can you fill us in on some of the copter breakages, battery situation, etc.

And again, you all did an amazing job running this session.

[aulvi]

Room Layout

Our session was held in the Boogie Barn, a fairly large and open space with a few posts and the remnants of a small hay loft. The barn was divided into three primary sections:

[ 4 tables ]

[ Flying Area ]

[ 3 tables ]

Each table was equipped with 5 chairs and two drones. In the Flying Area we had two groups of three chairs for teams who hate tables. We ended up with a max capacity of 16 drones and 41 campers.

Session Format

Each drone came with a printed copy of the AR-Drone module's readme, this was the primary reference used throughout the session. Unfortunately, when I printed the docs, the repl formatting was lost and that confused a few folks, but overall the docs were super useful. I brought 20 copies but should have brought closer to 40 to allow for more loss and damage.

As folks arrived we would ask that they self-organize into teams of 2-3, any late arrivals were usually guided towards an empty table. @nvcexploder and @nexxy opened each session with some words of advice, mostly around what NOT to do (ie, fly the drone into the sun). Aside from that, we mostly worked one-on-one to ensure folks were making progress and having fun.

Transition between formats was a bit rushed, we had to get folks to return their drones, collect abandoned drones, rotate batteries, assess/repair damage and re-set the tables/chairs.

I feel that our session duration was about right, folks had enough time to get their feet wet and decide if this was something they wished to pursue further.

Open Hack Time

We had a really solid turn-out during the open hack session. Providing an opportunity for folks to continue working after the initial session allowed for bigger thinking, and some of the demos that came out of this conference were really impressive! I hope those that gave a demo during the fireside will update this Issue with links to github.

Managing this session was pretty chill, folks were a lot more self-serve with batteries and repairs, allowing us to hang out more and learn how folks were solving problems. Additionally, some folks had fun just flying the drones around, the most popular option was either the iOS native app or WebFlight (http://eschnou.github.io/ardrone-webflight/)

Batteries

I believe Walmart Labs brought 6 drones (5 x AR Drone 2.0 and 1 x AR Drone 1.0), although we only used the 2.0 drones for the sessions.

Initially we laid out the 16 drones and set aside three for hot-spares, but eventually resorted to handing out as many drones as possible. Overall this worked well, but we did have to ask a team to give up an extra drone to allow another team to continue, but this was rare.

Battery rotation worked well, we got folks into the habit of running battery() to check the power level and swap batteries when they hit 30 percent. This reduced recharging time and also helped improve reliability, when the batteries get low the drones get weird.

During the fourth session we began to run a battery deficit, at one point we had less than 10 fully-charged spares, but diligent rotation helped ensure that we never ran out of juice.

Breakage

I believe this is a complete list of broken items, or at least a list of parts needed to repair all drones to flyable condition:

• 16 x Indoor Hulls (these are the foam bodies of the drone)
• 8 x Central Cross (the main X that the motors mount to)
• 2 x Blade Packs
• 1 x Gear Pack (really, we just need a tiny circlip)

Additionally, two drones were lost in trees during the open hack.

The Hull damage was somewhat anticipated, we were able to repair all damage with gaffers tape, no drones were taken out of rotation due to a broken Hull.

Central Cross damaage was far higher than expected. Almost all of this damage occured inside the barn, when the blades hit the ceiling the motors stop, causing the drone to plummet from the sky. Unfortunately, this type of repair is pretty time-consuming and failures in this area were not repairable between session. In other words, this kills the drone (at least for the day.

[mikeal]

If you had to estimate the repair time for a Central Cross, what would it be?

[tmpvar]

@mikeal depends on the break, I think superglue/epoxy would have been a great thing to have around. We patched with zip ties and duct tape with limited success. I wasn't around for all of the repairs, but i'd guess each one takes 20-30 minutes.

[aulvi]

I'd say 30-45min, you have to strip the whole thing down, reassemble and then test fly.

[nvcexploder]

Replacing with a new cross takes a bit longer, as all the motors have to be removed and inserted into the new cross, adding about 10-15 minutes.

On Mon, Jul 1, 2013 at 2:21 PM, aulvi notifications@github.com wrote:

I'd say 30-45min, you have to strip the whole thing down, reassemble and then test fly.

— Reply to this email directly or view it on GitHubhttps://github.com/mikeal/nodeconf2013/issues/21#issuecomment-20311296 .