It's that time of the year again. I'm not talking about the holiday season, though. I'm talking about the time of the year when everyone starts asking me about the next AVC. If you haven't heard of the AVC, it's where we gather a bunch of people with their homemade robots and challenge them to autonomously run around our course to win prizes. For 2015, the tentative date is Saturday, June 20th. Check out the recap from last year:
You can also read all about last year here. But enough about the past, let's look to the FUTURE! For 2015, we will be holding the AVC at our new building. With this new venue come a couple of changes to the event. First off, the ground course will remain largely the same, but will have a new short-cut feature and a couple of new surprises.
But the biggest change for the ground course will be further bonuses and incentives. Since the inception of AVC, GPS has been the go-to technology. However, over the years, it has become less and less challenging to plug in coordinates into your code and navigate using GPS. We continually hear "GPS makes it too easy." So, for this year, we want to offer an incentive to NOT use GPS. We are still finalizing the rules, but for 2015, you will be awarded additional points for not using GPS. This should level the playing field and allow slower entrants that don't rely on GPS to compete with the faster-moving entrants that DO rely on it.
Also, the course will have a shortcut with dangerous obstacles. Should you choose the shortcut, be prepared. This path will be shorter, but far more hazardous. The full ground course will be largely the same size as it has been in previous years. The full details are still classified, so stay tuned.
What about aerial? Well, there's some good news and some bad news. The good news is that we will have some sort of aerial competition next year. The bad news is that it will be limited to multi-rotors (quadcopters and such). The course will be confined (enclosed) and be a series of obstacles and tasks rather than a time-based course. We are still working out the final details, so stay tuned.
So, to recap, AVC 2015 will be held at our new headquarters on Saturday, June 20th. Every year AVC gets bigger and better and 2015 will be no exception. AVC is open to anyone (competing or otherwise) so we hope you can either compete or just show up and cheer on the other teams. Stay tuned for more information!
Great video. My favorite part is the song. ♫ That Jeep ♫ Kinda scares me. ♫ Oh that Jeep ♫ Kinda scares me.
Thank you. We really enjoyed making it up on the spot.
I would buy that song.
Sweet, it's not the same weekend as my wedding! I can attend both without painful explanation! See you all there!!
Any update on AVC 2015?
great timing, we just posted an update today.
Just want to clarify, as previous rules have never stated it before -- is there a period before the contest where we can freely drive around the ground track, collecting data, etc?
This would greatly facilitate vision-based navigation, for instance, especially if classifier training and mapping can be done overnight. But I would imagine even collecting GPS waypoints would be a lot easier that way.
In previous years, we've had the course setup the day before for practice runs, as well as during the morning before the event. We also allowed some runs in between heats, BUT, this year there will be no aerial competition, so the heats will be much closer together. Expect that we will have practice time the day before, and before the event starts, but not between heats.
Will there be any type of event this year for a UHS (unmanned hybrid vehicle system)? We are planning to compete this year but trying to debate one the direction to go with the one we built last year. http://diydrones.com/profiles/blogs/uav-and-uav-landing-platform-close-to-completion
Looks cool, but the ground course would be the only place for you to compete, and that would be just as a standard ground entry.
It is unfortunate that the Ariel competition had to change so much. Even with the changes, do fixed wing craft have to be banned? Also, will other non fixed wing craft be allowed, helicopters, blimps, etc.
not as of yet. for this year, we're scaling back the aerial until we can figure out a better way to go about it. we will gather feedback from this year and see what is in store for the next AVC. there has been talk about a separate event just for aerial.
Be careful with the difficulty. Sure, you may have experienced AVC six times, and want to do something different. However, this is still the first time for some people, and some of us, while we have entered more than once, are still struggling with the first turn. Go easy on those of us who are still newbies.
In other news, Project Yukari 2.1 will be there, still struggling for the first turn...
for sure. it will not be any more difficult than past years. however, there is the OPTION to make it more difficult, if you want to omit GPS and take the shortcut.
I am not too happy about the "GPS is the all evil" attitude. IMHO the real issue is that folks run commercial available off the shelf HW/SW packages. Perhaps a tweak should be instead to award constructors points, for HW and SW.
this is VERY difficult to check, without looking at everyone's code. It's not necessarily saying GPS is evil, but if you want to compete to win, your best design would include GPS. For the next year, this is no longer the case. It gives people additional options, if they still want to win.
Fully understood that it might be hard to police (what isn't). Given that there are so many ways to cheat to begin with one had to rely on a honor system anyway.
But given how many people still struggle with the first turn (using their own designs) ...
While I'm somewhat disappointed in the change to the Aerial competition, I completely understand. At AVC 2014, I was in the crowd very close to where the quad went out of control and the helicopter crashed and burned (literally). Safety first!
Looking forward to a great AVC 2015!
I'm really excited about the bonus for not using GPS! It's a little more fun to pretend you are building something to operate in an environment where GPS isn't available... Luna, Mars, Post-Apocalyptic Earth, etc...
I'm actually pretty excited about it too. We're still working out the point structure, but in essence, if you have a vehicle that can do the course in one minute using GPS, you'd get about the same score as someone that takes the full 5 minutes, but doesn't use GPS.
Would you please clarify this comment ? In previous year(s) there were 300 points available if you finished under 5 minutes (essentially 1 point for every second under 5 minutes). For 2015 it seems you imply that the GPS using class would get 5 points for every second under 1 minute (i.e. compressed time scale). Do I read this correct ?
There are no gps or non-gps classes. there will be the same classes as last year. but if you don't use GPS, you just get a bonus of (still undecided) 200 points. 200 points would be equal to roughly 3 minutes and 30 seconds. So, it effectively shaves off that time. Does that make sense?
Ah, got it. So a added bonus for no GPS use, but the same time points as before. Makes sense.
GPS, does that mean GLONASS is ok ?
how about "no reliance on external positioning data derived from satellite or ground station". so, yeah, just on-board sensors would qualify as 'no-gps'.
smile Just a lame attempt for a bit of humor on my side ;-) Makes all perfect sense now.
One more detail. The to be decided bonus, is it per corner made, or are 200 (or 150, or 100) points granted after getting around the course ?
The rules aren't finalized, so this is all in theory at this point.
the points would be granted after the course is completed, it much the same way the time points are done. So, just not using GPS doesn't magically grant your run 200 points.
I'm curious if the GPS vs No GPS decision is made on a per-vehicle basis or a per-run basis? If I try my first run w/o GPS and fail miserably, can I put a GPS in for future runs?
I'm also contemplating 2 nearly identical vehicles, one w/ GPS one without. Would this be ok?
Sorry for the delay in answering, I was out of town.
We'd rather you not modify the vehicle mid-race. Each entry should be the same vehicle between races. Once we start allowing people to turn on/off GPS, there's little difference between swapping out sensors, the chassis, etc, and then it's hard to tell what vehicle was actually entered. Pick a strategy and stick with that. OR, register two variants like you said.
Good to know. Thanks. Now, we just need those rules to pick strategies. :)
Is there an ETA for when full details will be announced and we can start designing our new bots?
We really hope to have it nailed down in early January. The ground course will be easier, but the aerial might take a bit more time to iron out the details. We will have an update in the next couple of weeks.
The team is on the edges of their seats waiting for the next update. :) Take your time, though, we're just excited for this year's event.
we're working on finalizing the details and should have a blog post in the next week or two, stay tuned!
Is the ground course going to be around the parking lot... or around the building? IE can we range find off of a straight wall...
It will be like last year, around the parking lot. our building is a lot bigger, so the course would be almost double the distance if we did it around the building, and there is a sidewalk along one side, so it would have to transition into a street, etc.
I'm curious about the "dangerous obstacles" in the ground course... does this mean they will be actively trying to inflict damage on the competing robots?
No. We're not going to have hammers, minefields, or anything like that. But we may have something along the lines of an active barrier.
Will the "active" part be an employee?
nope. it will be mechanical.
That sounds pretty cool. Thanks for clarifying!
Regarding the aerial course: By enclosed, do you mean indoors?
we're not entirely sure. It MIGHT be indoors, but most likely it will be an enclosed (either with nets, or a more permanent physical barrier like plexiglass) arena outside.
I was just worried about GPS. Hmmm, now that I think about it, I guess we don't really need GPS signal since we are avoiding obstacles. We'll need computer vision and other sensors mainly.
Thanks for the freaky fast reply!
yep, this would be much more reliant on computer vision, you're right on!
I'm the only one here, so all I can do is reply to customer comments.
:( I assumed (we all know what happens when we do that) that It would be the same and spent the last couple months at my school working on our red balloon find and destroy system
So sorry about that. It was a difficult decision, but the aerial event gets more and more dangerous every year and we needed to make some changes.
I had a guy donate some parachute material for my ornithopter. I guess thats out. If it hovers can i compete with the quads?
no worries, we learned a lot and hopefully can apply it to this years challenge :)
No pond(s)???
There will be no water obstacles at AVC 2015, since there is no water near the new SparkFun building in Niwot. Yay for no water!! Or boo, whatever's your thing. ;)