Looking for some projects

How about you have a go at an AUV?

"Michael"

Although it would be nice to see a sail powered (motion I mean) AGV. Seems like he already has the platform.

robot, but I've done all I want for the Sail Car from help with development through making all the tooling for their production.

I'm looking for simple metal, wood or plastic details to help you robot designers and builders in exchange for just sharing ideas so I can learn more about the electronics and software part of the equation. I'm pretty good at the mechanical side...

Wayne

"Wayne Lundberg"

Wayne,

Perhaps we can share some information or even put something together that will be pleasant to work on. Let me first introduce myself. I am a software engineer with about 12 years of professional experience. I work here in San Diego on a software/hardware manufacturer

and I'm taking my master's in computer science degree through SDSU My thesis is inspired on the DARPA Grand Challenge. Unfortunately SDSU didn't have the budget to enter the competition with a full size car, instead they asked me to create a low budget robot ($3K to $4K) based on a electric RC car platform. The challenges of autonomous navigation are very similar no matter the platform you're using. Obviously we cannot enter the DARPA GC with an RC car. We are still asking for a bigger budget to buy a platform and equipment that would be able to traverse 200 miles of desert terrain, but with current state budget cuts, I doubt that will be the case. Here's a link that shows my little robot in early phases of development (mechanical phase is finished, I'm working on electronics right now and I'm expecting to start the AI software early next year). My idea then is to create my own robot. I don't have the money to buy an offroad car only for that purpose, but I still think it is possible to create a cheap platform able to cross the off road miles. My idea for a platform would be a home made vehicle made out of ATV parts.

I saw that you also live in Chula Vista (I live in Eastlake), so we have a geographical advantage. It seems that your skills complement mine. In a scale from 1 to 5, I would rate myself 4 or 5 for software, 3 for electronics and 2 for mechanics. If you think we could do something together, let's talk sometime and I could show you what I have done already and what are my plans for a DARPA GC entry. (if nobody wins this year, next year the prize will be $3 million dollars!)

Cheers,

Padu

---snip for brevity---

Sounds like an interesting challenge! First off, I suggest we go to the junk yard up by Brown field and chat with some of the salvage operators. I'm sure we can pick up a power unit without body and totally unsalable to the public for use on highways but ideal for a desert and off-road vehicle as you propose for a couple hundred dollars that I'm game to put into play. But first of all, I invite you (and anybody in this newsgroup that could make it) to a backyard Tequila sipping or Michelada tasting, note-taking session to plan the plan... I may even put in some guacamole and chips for added joy. Heck, we may be able to get a place to build the device at the Southwestern College Auto Repair shop! (They could use the publicity!)

Anybody care to join us?

Wayne

It would be really, really cool if you and Padu could ride up together to the next San Diego Robotics Society meeting, which is held the first Saturday of every month. We have it in a member's garage, which we have full use of. BYOB, but someone usually comes with food and treats of some type. The meetings are held in San Marcos; map is available from the

Web site. Most of the members are into larger robots, though to my knowledge no one is currently making an offroad AV. I think you might find some other interested parties, as well as people who'd like some things made for their projects.

-- Gordon

"Gordon McComb"

I missed 3 meetings already, I'll make sure I won't miss the next one.

Padu

Give me a call Padu the day before and we can car pool the trip. Maybe get Rick to join us also?

Wayne 421 6728

805 & Olympic Parkway

Hi all, My name is Reuben, I do production machines and automation for a fibre composite manufacturer

I'd love to join you in a $3M winning venture...but I live on the other side of the rock :)

BTW, we have a bridge on your side, and another one on the floor now going to NY state

Cheap toyota truck at the junkyard: $300 gear motor to control steering : $100 solenoids for gas, brake, clutch : $300

gear motor + 2 solenoids for switching gears: $400

I am making high estimates. Most of this stuff can be had for cheaper or free in the classified ads of your local paper (the RECYCLER is an excellent source)

That's $1100 for the vehicle + actuators. need some sort of feedback from the sensors.

computer could be $150 PC. USB cameras are $10.00 apiece.

wiring: $100. I/O card for the PC $150

$410 added to the cost.

You need people to build this thing, not money. If anyone in the San Francisco/Sacramento area wants to start a project, I'll get a toyota truck for the project. Also, the gear motors, PC, and USB cams.

Rich

Software which can reliably drive a truck at 20mph over 100 miles of rough terrain using two cheap USB cameras as sensors: priceless!

Tim

Yep... not doubt about it. Who do you think you are seeing when driving down around Chula-Juana hiding cheap wine in an old paper sack because life and technology has left him behind in the garbage heap of old Toyotas and washing machine motors, vacuum tube radios and manual typewriters? And even with a fully functioning robot, PLCs and stepper motors spilling out of his garage, nothing seems to matter because software developers are nowhere in sight... maybe trying to stay alive by working twice as hard because of the Indian onslaught?

I am working on other projects right now, but I'd bet a large sum of money that I could get a truck to go 100 miles over rough terrain in about 1 year, doing all of the programming myself.

Or more precisely, $2,000,000:

-Brian

Since you are making pointless "all you have to do ..." vapor bets, I can make a truck cross over the rocky mountains at 550 mph without any programming. ;-)

One thing that got me very motivated to enter the GC is that not more than 5 miles separate the $2M team from the $2000 team on last year's results. As someone else said here earlier, the key is in software, not hardware. Think about it, a dog is able to dodge obstacles while running at very high speeds, and it doesn't carry any ladar, only a pair of cameras and a gyroscope, but a very refined software package.

Padu

But a dog has a computer unlike any you could possibly buy off the shelf. That makes a big difference.

My dog only has an old Zaurus. It's all he could afford.

Check out our Darpa GC entry here:

Software is indeed the key, but it is definitely a weakest link issue. Poor hardware and sensors can make the software's job a whole lot harder or even impossible, of course. We've considered the "brain in a jar" technique, but nobody on the team is willing to volunteer :-)

The DARPA rules specifically state "no animials are permitted onboard"

- not sure if a brain in a jar would apply or not.

The NQE should be exciting. We are expecting tough competition and are working hard to be one of the 20 finalists.

-Brian