[EMS Discuss] Red Bull Meeting Saturday
Fri Apr 12 10:24:00 PDT 2013
Don’t know what you were wanting for the chassis of the gun, but something old school might be cool..
From: EMAIL HIDDEN [mailto:EMAIL HIDDEN] On Behalf Of Benjamin Hallert
Sent: Friday, April 12, 2013 8:43 AM
To: Rick Osgood
Subject: Re: [EMS Discuss] Red Bull Meeting Saturday
Good point re walking off. If its a Raspberry Pi in the gun, it would be tragic if it didn't play some sort of cool shooting sound file on trigger pull. The sound hardware is there, it needs only a speaker.
Ooh, and a solenoid that fires for 'kick'!
On Fri, Apr 12, 2013 at 8:36 AM, Rick Osgood <EMAIL HIDDEN <mailto:EMAIL HIDDEN> > wrote:
The wiimote idea is interesting and no one had thought of that last night. I think Kassie is planning on looking into OpenCV today to see what kind of performance she can get with it on a RasPi. For communications from the guns to the Bullduino, anything can go really. Last night we had talked about using zigbee wireless modules or even just using a physical cable with serial or something going back and forth. We were thinking it would be best to have the guns tethered anyway so drunk people don't try to walk off with them. There are probably a lot of ways to do that though so I guess whoever is most interested in it can research the best way to do it.
On Fri, Apr 12, 2013 at 6:57 AM, Ben Hallert <EMAIL HIDDEN <mailto:EMAIL HIDDEN> > wrote:
Regarding target acquisition from the guns, I have two areas of research to suggest:
1. Wiimote - the Wiimote has a camera that can track up to four 'blobs' in hardware and can interface through I2C if I remember correctly; this could reduce the CPU requirement, especially if it's determining if the blob is centered. Putting a scope on it might increase range and improve accuracy too.
2. OpenCV is a vision library that may be able to assist as well if we want to go with webcams in general. If we can illuminate the target locally with infrared (this would be part of the Wiimote strategy too) it could potentially cut down on false positives and help us make sure the gun is detecting that it's actually pointed directly at the target.
This is going to be great! WiFi dongles in the guns, I assume, for range? And an AP on the target for both guns to connect to?
On Thu, Apr 11, 2013 at 10:41 PM, Rick Osgood <EMAIL HIDDEN <mailto:EMAIL HIDDEN> > wrote:
After our meeting today we decided we should meet again this weekend to hopefully allow other people to get involved who couldn't make it tonight. We were thinking Saturday morning at 10:00AM at the shop. Hopefully that will work for some people who couldn't make it tonight. At the meeting we want to plan out the design in more detail and also start distributing work out to people. Now a re-cap of what we did tonight:
People present tonight: Rick, Kassie, Michelle, Cord, Marie, Taper, Hovis.
First we went through all the ideas to whittle them down. Ultimately, we all came to the conclusion that we liked the "laser target shooting" idea the best, and it seemed like one of the more feasible ideas too. Plus, it turns out that Kassie was already working on a similar idea for a game and has some circuits already working for her own game that we can use in this project to help the project move faster.
Then we did some brainstorming to get some general ideas about how it could be designed and how it could work. This got us all really excited for this idea so it seems like this is the one we would like to stick with. Here's a general concept we worked up tonight as something to try:
There will be two laser guns. Each gun will have a RasPi inside with a camera. The RasPi will do image processing to detect when the player is pointing the gun at the target and pulls the trigger. It will then transmit hit/miss/etc to the Arduino in the target wall. They will also have a laser inside so people can see where they are shooting and to add more LIGHT.
The target wall will be all lit up and controlled by the Arduino. The targets will be rotating flat targets. As a simple example, picture a playing card. In it's normal state it will be sideways so you can't see either side in order to shoot it. The target will randomly rotate 90 degrees one way or the other to show one of the two faces. One side is for one player to shoot, one side is for the other player. When someone hits a target there will be a physical bell that dings. We liked the idea of having things physically moving because it gives that feel of a classic carnival game and it's more interesting to look at.
The wall will be all lit up in LED strips. The color of the LEDs is determined somehow by whoever is in the lead. So whoever is taking control of the points, takes control over more LEDs on the wall.
We want this whole thing built into a bicycle trailer so we can haul it around downtown and let people play it.
We also talked about making an overall theme of "cowboy robots" or something similar (geek meets vintage-arcade-shooter).
We had a bunch of other ideas as icing on this cake but I won't mention them here because they aren't essential to the build. Kassie has a bunch of free time tomorrow so she's already looking into using image recognition with the RasPi to see if it will even work for us. If not, we'll have to do target sensing another way. Taper is going to start looking into the RedBull arduino library to start getting a handle on that.
Anyone who wants to be involved in this project, please try and show up Saturday morning so we can nail down the details and assign tasks. If you can't make it, let us know what level of involvement you can commit too over the next three weeks and let us know what you are most interested in so we can figure out something for you to work on.
Discuss mailing list
EMAIL HIDDEN <mailto:EMAIL HIDDEN>
-------------- next part --------------
An HTML attachment was scrubbed...
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 140489 bytes
Desc: not available
More information about the Discuss