|
#91
|
|||
|
|||
|
Hi,
I think it's a little early to start a web site, I know we will soon out grow the Delphi Forum, and be come too much of a burden on these Wonderful people, But I hope they put up with us a little longer, This is the best place to contact FIRST people....... I have ask Brandon M. to start a New Thread using the two proposal posts And this maybe this one(?) Titled "The Grand FIRST Team" (I did not get any feedback, so I hope you all like the name. And I nominate it for the Official TEAM Name....... Any others? 2nds?) I asked for this change To generate New Interest and ease of uses I am glad the question of organization has been addressed early, (Thanks, Paul) This is one of the things that will make or break this team, that and Communication are VERY, VERY IMPORTANT I think its also a little early to be picking people for posts, we have not even agreed to do this yet, Let-alone set the team structure. I will say that Matt is in a very good location for the So West just as Greg is centrally located for the net. Just PLEASE, EVERYONE look inside your-selves and ask, do you want to do this, And.......... if you have time for this, Some of these JOBS are INTENSE And don't be afraid to ask for help if you become swamped, Thats what a TEAM is for! One last thing this time, Its still early but I would like anyone interested in this endeavor to contact me,Personally with real names, locations, team #, Short resume, and permission to Use / Post Your Name in conjunction with this contest to rally sponsors and gain interest Keep posting, your thinking is IMPORTANT! Geo. P.S. think I found a GPS expert! P.S.S. If I have not mentioned any one, Don't feel "left out" Please contact me. |
|
#92
|
|||||
|
|||||
|
I'm in proboly one of the least geographically appealing locations for this, but if anyone on the easy coast thinks of anyway we can meet occasioanlly, let me know! I'd love to help but I think the real key to this whole thing, truthfully, is money. With money, the two debated main points of the competiton here, Navigation and Machine durability/efficiency, will both follow. Is there anyone out there who's amazing at finding sponors and knows somebody/someplace willing to give a couple hundred grand (because, realistcally, that's how much this kinda stuff would take).
Now I've gotten that outta the way, my thoughts on the machine. I think that the standard commercial vehicle drivetrain would be needed, so it's really kinda pointless to reengineer something as complex as a motor system. However, it's really also pointless to take a jeep and retrofit it with a servo control system to control steering etc. There would have to be a happy medium. If money weren't an issue, I'd say build your own Chasis and rip the engine system out of something really efficient, like the new VW's Diesel engines. Build a strong, durable, aerodynamic, lightweight chasis. You don't need to take passengers, so that cuts out most of the inside of car. Now you've got a chasis thats about 3-5 feet high at most, somewhere around 4 feet off the ground with huge beefy tires. Control system-wise, I've seen a lotta research going into vision detection. I think that would be great for line tracking while on the road. Once offroad, that system would be disabled and switch into an alternate system for object avoidance. [Object avoidance would be activated during line tracking, but would not be the primary guidance]. I'm also thinking, if you were to get a map of the overall area, and plot the course using way points, AND indicating alternate routes if the primary is unavaible, you'd be in good shape. Walk up to the machine in that two hour period, pull up the map of LA to LV on a tablet PC, and just start plotting where its gotta be and when. If it discovers through obstacle avoidance that that routes not available, switch to secondary or triary waypoints and continue. Waypoints would be established and checked against GPS and distance. Your waypoints would have distances according to the map (as long as they're accurate) and you could check yourself between the two, confirming your in the right place. Well, thats my thoughts... |
|
#93
|
||||
|
||||
|
This intrigues me greatly.....I'd like to offer my assistance wherever necessary. When taking on such a task, there are so many variables to consider, you guys are doing a fine job touching many of these variables, but having read the posts in this thread, I thought I'd offer some of my input.
What about traction/wheel durability? Say your'e machine had to take on a steep grade, you could shift gears...but without the finese of a human foot on a gas pedal, you're wheels are going to slip. Also...What if you unfourtantly ran over a sharp rock or something along the lines of that, and punctured a tire? Goodyear runflats would be the simple solution...but I don't think they have a baja eddition of that product line. |
|
#94
|
|||
|
|||
|
Quote:
For they type of wheels and tires, I'm thinking some 33"+ Super Swampers or Boggers or possibly Pro-Comp X-Terrains. Any of those should hold up fine and have excellent traction. For wheels, you definitely need beadlocks. There is a world class off-road shop in Burbank, California called ORU, (Off Road Unlimited) who can do any custom suspension beefing-up or lifting you want. Also down the street from them is a 4 Wheel Parts store. |
|
#95
|
||||
|
||||
|
Controlling acents and decents would be yet another interesting challenege to overcome. Take an incline at an angle, and you might roll right down the hill, or stay on the incline too long, and the enginee could flood....Take a decline to fast, and you might lose controll and careening down the hill, only to smash your'e creation's front end when you hit the ground.
|
|
#96
|
|||
|
|||
|
Quote:
|
|
#97
|
|||||
|
|||||
|
Quote:
I think maybe for something along those lines would be prevented through software. If the software sees that it's coming up to something it's not going to travel across well, it's localized navigation system would kick in and find an alternate route. This could be created using obsticale avoidance (as any sharp incline would rise up to a level where the onboard sensor suite could detect) and through something along the lines of a gyro. The gyro could give the software feedback, letting it know that the car is at too much of an angle, and needs to back off and find another way. |
|
#98
|
||||
|
||||
|
It could be too little too late to back up and find another way if it's already going down, or up something.....What would happen if somehow, the truck managed to high center itself on something? Say it's approaching a rock, the onboard sensors detect that the rock is x inches high, and the program knows that the suspension has y bound....without some way of 3d imaging, the rock could appear as x hieght from one angle, or another variable from another...so say the vechicle decides to go for it, and attempts to go over the rock. Midway through the transition over the rock, the wheelbase proves itself to be too long and the suspension doesn't have a large enough bound...causing the truck to become stuck...
|
|
#99
|
||||
|
||||
|
The idea of having a wheels too far apart causing the autobot to get stuck reminds of issues we had thinking about this year"s FIRST game involving the ramp. There will probably be ditches and slight bumps that will be large enough to cause problems for 4 wheels, but if we"re going to build the chassis ourselves and just use a premade motor, why not use six wheels instead of four? This would help us clear ditches deep enough to get the front wheel stuck in, rocks or fallen trees big enough to catch the underside of the autobot, or patches of mud that compromise the traction of two or three of the wheels (and it would look alot cooler than anything with only 4 wheels).
As to object aversion, there are ultrasonic sensors that come with a servo and a small circuitboard that check distances between the sensor and surrounding objects. Coupling several of these at different levels with the velocity given by the GPS unit, a usable map of distances and heights between the autobot and its surroundings could be created which could be used for plotting the course. (And just for fun and appeal to The DARPA, I think we should put a turret on the top with a pyrometric sensor and a digital camera that snaps pictures of everything living thing and autobot it comes across and pitch it as a "target recognition and tracking device") |
|
#100
|
|||||
|
|||||
|
Quote:
|
|
#101
|
|||||
|
|||||
|
Quote:
the difficulty with steering could be solved (at least) 3 ways: [how old tanks solved it] the seperate treads on each side could be hooked up to seperatea clutches and brakes, and you could use these to power and de-power(is that a word?) the treads independantly [how new tanks solved it] the engine could feed turn a generator that would power an electric motor for each tread finally what if you where to use two smaller engines instead of a large one? i think for simplicity and weight, the first is the most suitable idea (that i know of) anyways, just figured i'd fling some ideas at ya'll... <edit> what if instead of clutches you did something with CVT's...</edit> |
|
#102
|
|||
|
|||
|
Hi again,
I think everyone needs to thank Brandon M. & Delphi for Letting Us use these forums. Please don't forget them in your next posts. Dan L.(247) has an interesting point about declines/inclines (something I missed) add to that holes, (how deep?) washes, (How deep/steep/wide and whats the other side look like?) Water (how deep/wide/whats the bottom look like?) Rocks (how big/high/whats behind it?) these are the type of Questions We (and the Truck) have to ask. AND ANSWER! The sooner the better for the Team With in Seconds/ at the most, Minutes for the Truck! ( Plus "where am I??", "where do I have to go?", And how do I get there?") Paul Mentioned ultrasonic sensors, Banner Engineering makes line of measurement sensors with a 10' range could we gang a bank of these to "draw" a picture of the area? Something On this line would go a ways in answering some of the questions. Dose anyone have any experience in this field? How about a Mentor or (gasp!) parent? Come-on someone knows about SONAR ARRAYS Lets Pick some BRAINS! One of our biggest ADVANTAGES over other teams is the network that we can create thou our participation in FIRST! if you notice I keep referring to the vehicle as a "Truck" all thou other forms may be advantageous(custom chassis,tank treads etc.) the most practical thing is a Truck, Being out in the dessert, there is a lot of time spent building sand rails, rock hoppers, trucks, bikes, and other toys. I will post some pictures of TWO recently completed projects soon. I am under no delusions about how much work a project like this is. or how much money it takes. (a Good rail is about $35,000 from the ground up.) And for this contest we will in all likelihood build 2 trucks One just for testing (it will get Hammered!) and one for the race. I am hopping to get the trucks donated, But we have to show the Sponsors Something! (See earlier posts) Please visit this web site to get an idea of the machine shop backing we have available, Please go to http://www.laron.com (sorry, I don't know how to make a link) Please keep adding your IDEAS Please keep posting Please ask people that might like to help this program please recruit Geo. P.S. As for meetings, how about FIRST events? If we spend the time (WISELY) from now to the first race getting organized, we could meet at nationals, I will try to bring a vi-do of the run, And this gives us about a year to build & test for 2005. THINK, and have FUN! |
|
#103
|
||||
|
||||
|
Truck, eh?
For those of you who read Wired there was an article in the last issue (Stopping Loose Nukes) there was an article about Robert Full and his research into legged robots. He is a professor at Berkley (good location for helping build this) and might be willing to join us on this team (this is also a shout-out to any members of team 225 to recruit his help). I believe that a walking robot would be much more agile than a truck, but it may not be as fast or as able to take a beating and keep going. We need to get an idea as to whether or not something that agile would be able to make up for the lost time by being able to pass through rough areas with no loss of maximun speed. I've never had a chance to hold a good scientific investigation on the geography of the rougher areas in California and Nevada. Any insights into this dilemma? |
|
#104
|
|||
|
|||
|
Quote:
No, seriously - a Jeep.Last edited by sanddrag : 28-05-2003 at 23:28. |
|
#105
|
|||
|
|||
|
Truck
I still think Truck over Jeep
1) Longer wheel base / more stability 2) More cargo area 3) I might have sponsor (if you have someone willing to donate, I have no objections to Jeep other than short wheel base, I have had a Toyota LC (FJ40) for 23 years and if your not careful your rubber side up with in no time. You want a longer wheel base, much more forgiving! Geo. |
![]() |
| Thread Tools | |
| Display Modes | Rate This Thread |
|
|