Comment

Comment by JB on March 12, 2013 at 5:24am

Sorry Guys but I've already committed myself! ;-) If not for the OBC then for a surveillance platform for commercial use. The OBC will be a bonus.

I'm pretty confident we can achieve it if most of us chip in. I think the min. we need is about 5 people atm.

I'll bring a gant chart with dates and deliverables for completion by the team. I have also already started to do some system breakdowns.

I'll pass the web thing on to my friend and try to get him interested as well. The more the merrier.

In the mean time I'll make some more notes on the report Hai posted. Does anyone have links to some more?

@Stephan Can you post a link or email me the last OBC rules please? Thx.

Comment by Stephen Gloor on March 11, 2013 at 11:59pm

@Hai Tran - "What's going to make us successful or not is project managing it so that the deliverables are met. "

Never a truer word was spoken.  Will see you all on Sunday

Moderator Comment by Hai Tran on March 11, 2013 at 11:56pm

@JB pass on managing the forum/site. 

Lets meet up at 10am this Sunday at Coptercam HQ and talk about building the team and making sure we have the resources and commitment we need.  We should take a read of the previous year's rules and format, plus an publications from teams are part of their deliverable.

Let's not jump straight into solutions model.  I agree that we should break the problem into manageable blocks/systems, but it's pointless getting into the finer details of scanning, bomb calculations at this stage. What's going to make us successful or not is project managing it so that the deliverables are met. 

Comment by JB on March 11, 2013 at 9:49pm

@Hai

Did you want to manage the forum/site? A friend of mine who does websites as a job would otherwise be able to do it for us, and he might even be interested in doing some web gui for us as well. In the beginning I thought we could setup a google group otherwise.

@Stephen

I didn't want to come on too strong...but sadly that's in my nature! I like this stuff!;)

That's good news with 75minutes each, at least we won't have to contend with other RF systems. Do they employ "radio-silence" from all non-active participants? If so we might need to do some off-site testing prior to launch.

I have already made a group so email me for an invite on jeff blogg at g mail dot com, all written together of course, as a trial. If those involved send an email there I'll post back an invite to join. Note: I'm only doing this for test purposes atm to see if it will do what we want/need. We can then migrate to a "real " one. I think to get full functionality a gmail account for registration would be best, that way we can use google drive to sync docs etc. one good thing with google groups is that we can do invite only, so everything can remain confidential until we are ready or willing to release information. BTW can you upload the previous OBC rules once your on the group, or email it there?

@ James

Image stabilization is required especially so if we are going to use telephoto optics. The downlink might also need stabilization tracking depending on the system we choose in the end.

The RFD900 radios have already made it on the list. We'll need to do some serious RF testing for cross-interference from the other RF sources. We might need to keep to 2-3 RF bands outside of each others harmonic bands. Like 433Mhz RC, 900Mhz Comms and 5.8GHz downlink. The heavy data bandwidth ones need to be at higher frequencies. Also antennas and placement are nearly more important than RF power output, but it's always good to have both! ;)

Hopefully I'll see all of you on google groups soon.

Regards

JB

Comment by Stephen Gloor on March 11, 2013 at 8:04pm

@James - Agree on the stabilized mount - it is pretty hard to keep the plane perfectly stable.

Comment by Stephen Gloor on March 11, 2013 at 7:52pm

Everyone gets their own 75 minutes actually.  There is setting up time.  I think a flight controller is a good idea and the pilot is the ONLY person that sends commands to the UAV.  Once we have the coordinates that needs to be given to the organisers for checking.  Only if we are within 10m are we allowed to drop.  I think we will give it to them in all possible systems to avoid the WGS66/WGS84 stuff up that cost CanberraUAV a drop.

Comment by Stephen Gloor on March 11, 2013 at 7:49pm

@JB - Sorry for appearing that I corrected you - that was not my intention as your ideas seem like gold and you have the required knowledge that I completely lack.

These are all the details we need to work out and enthusiasm is a really good thing. Maybe I have been in IT too long and seem to many great ideas not work out in practice.  I think I have been in IT to long :-(

On the geofencing I would really like that to be something separate like a GPS tracker with geofence capablity.  If the coders among us can 'borrow' the APM geofence code and get it running on a watchdog board then we could have the autopilot geofence set just inside the real one so we get a warning if we get close and before the plane crashes.

I have a copy of the 2012 rules that I can send out.

@Hai Tran - Completely and totally agree on the payload delivery.  This is the big missing thing on all the previous teams as they only seem to have a vague idea of dropping a parachute.  I think the payload should be a bomb with fins that we drop from a dive.  A steel water bottle with fooze ball fins attached should track straight.  Will also need to know the wind direction so we dive directly into the wind.  This is of course unless we can have a smart bomb with terminal GPS guidance.

Comment by James masterman on March 11, 2013 at 5:36pm

A stabilised camera mount is probably a good idea. Even in light winds roll stabilisation helps.

For comms, I've been using the RFD900 radios. These are a high power version of the 3DR radios. Should be fine for long range telemetry.

Google maps will fit into the solution somewhere, but its worth noting that their imagery is often not accurately positioned.

Comment by JB on March 11, 2013 at 9:32am

Hey Stephen

Sorry for rambling on with the imaging stuff...once I get my teeth into something I try to get a good taste of what it's about before I quit. The imaging shouldn't be  a problem, but it will be the core bit of work to nut out. As I said a couple of posts ago, what we are building is essentially a flying scanner, we need to get the dpi the right size to read the letters, so that we can find Joe.

I completely agree that the pilot needs his own OSD, GSC, camera and downlink. This can be done "conventionally" via a analog link and 3DR or the like comms. The geofencing also needs some consideration to avoid having to ditch the airframe. I'm also concerned that the SAR area border is the same as the geofence ditch, if so we'll need to have sideways "looking" ability to stay away from the geofence.

On weather: high wind speeds will present more than one problem. 1st unstable aircraft = unusable camera 2nd non-symmetrical passes with high overlap 3rd range issues 4th difficult approaches for delivery 

Cloud cover under 200m can result in a complete strategy failure if we choose a +ALT400m plan, so we will also need to work out contingencies as well to cover all variables. These things will have to form part of the testing procedures...ie some storm chasing.

I agree that we will have sub-teams inside of the group. 

On staffing: Is their a limit to the crew allowed to be involved in OBC?

Hai I think a Payload Specialist would be required, who can also handle any service ground work like battery/system maintenance etc. and of course getting the delivery on target.

The two most important guys who aren't allowed to multitask at the event is imaging and guidance. Piloting and comms are really only redundancy positions if something fails with the other two.  A couple sets of eyes on the targeting monitors will probably also be required.The comms need to be sorted way before we get to OBC, otherwise the whole effort is a bust. We can limp on with manual piloting and human live-view of video or stills...but nothing will work without comms and the whole thing will come down in flames.

BTW are we competing at the same time at the event with the other UAV's or does everyone get their own 60 minutes time slot? If we're together the RF bands are going to be crammed full with us all up there at the same time. Anyone going to play flight controller? Has anyone found a copy of the old OBC rules? We need more info!

One last thing: we might be able to use Google map maker to overlay images in near realtime, and use that as the target search interface, instead of a clumsy photo editor. At least with map maker the GPS data is a given, and we can then get permission for delivery. I'm hoping the water bottle is completely full...

Moderator Comment by Hai Tran on March 11, 2013 at 8:24am

One more role Stephen is payload delivery, which seems to get entrants every year, I don't believe that any team from the main competition has yet been able to deliver the water at much accuracy.  I believe Canberra UAV almost did it, but they payload delivery system failed somehow.

It's 2:20am in Sydney, I was on the 5:30am flight out of Perth, time for bed before the flight back in the morning.

• ‹ Previous
• 1
• 2
• 3
• 4
• 5
• …
• 32
• Next ›
• Page