Archive for March, 2004

Wait is on at Slash X Cafe

Saturday, March 13th, 2004

We are at Slash X Cafe. The local time is 5:15am. All the teams have received the waypoint file, the route that the robot should take. We have projected the route on satellite imagery of the area, provided to us by Brent Yantis of NASA Regional Application Center. If one were to assess the difficulty of the route from the imagery, the route looks straightforward. It appears to travel around the Stoddard Valley OHV Area, then follow powerline road into Barstow Area, go through a section of Needles, and then head on straight towards Primm following a trail that was once used for Barstow to Las Vegas dirt bike race. The route appears to be best for the nimble. Two days ago we attempted to add NTP capability which is necessary for gaining speed, We were unsuccessful. Hence, our current plan is to run at a snail’s pace. Most people do not expect anyone to complete the whole route. Given that, our goal is to be the last person standing. The opening ceremonies start at 6:00am (local time) and the race begins at 6:30am. The DARPA Grand Challenge is likely to be an important event in the history of technology. We feel proud to be part of it.

Arun Lakhotia

Schedule for the BIG day

Saturday, March 13th, 2004

Folks, the BIG day has arrived. For our fans, here is the schedule for the day. First the time zone. We are in Pacific time zone, two hours behind Central Zone (Louisiana time). When it is 3AM here, it is 5 AM in Louisiana. The team will wake up at 1AM (yes, that’s AM, an hour past midnight). Pack bags, and head out at 2AM. Well, at least the ‘bot’-handlers will be out by 2AM. Others will follow within the next half hour. We expect to arrive at Slash X Cafe in Stoddard Valley OHV Area around 2:45 AM. We must be there by 3:20AM. However, we do not want to take chances with car trouble, traffic problems, and the likes. At 3:20AM DARPA folks will give the first six teams the RDDF file. RDDF stands for Route Data Definition file. The file will contain the route DARPA wants our bot to take. The route will consist of a sequence of ‘way points,’ given in latitude/longitude form. Also included in the route file will be the width of the track between two way points and the maximum speed permissible along the track. The next six teams, which includes CajunBot, will get the same information half an hour later. The opening ceremony will start at 6:00am. At 6:30am, the first bot from the Red Team of CMU, will leave the chute. The next five bots will leave at five minutes interval. There are only six chutes at the opening line. The next six bots will then load up the chutes. Ready to leave at 7:00am. From there on the schedule depends on the performance of individual bots. Most DARPA folks expect that the field will get narrowed down to 10 within 3 miles, and then reduced to 3 within 20 miles. The route is expected to be tough.

Arun Lakhotia

Track CajunBot during the race via internet!

Saturday, March 13th, 2004

You can track the race via this website tomorrow! -aDam

Tracking the Race

aDam Dupre’, adamdupre12@yahoo.com

CajunBot in Slash X Cafe near Stoddard Valley OHV Area

Friday, March 12th, 2004

CajunBot is now at Slash X Cafe, the starting point of the challenge. The cafe is at the edge of Stoddard Valley OHV (off highway vehicle) area. The place is around 4,000 feet above sea level. I had always associated high elevation to cooler weather. This place has questioned that belief. This place is hot. It is hot and dry. Even though the place is surrounded by snowcapped mountains, the soil here is dry. As I drive around the town I am reminded of Rajasthan, my home state in India. The soil is dry and lose. There are green shrubs all over the place, but barely any large trees. Within an hour in the open, your face gets covered with a thin layer of dirt. The houses too have earthen tones, just like in Rajasthan. In their free times, adults and kids in this area, take out their dirt bikes and travel the OHV areas. They take pride in not taking the beaten path. That also explains the numerous ATV trailes, which look very similar to the mud roads of Rajasthan. The race is expected to be on such ATV trails. I am using the word ‘expected’ because we still do not know the route. It will be given to us two hours before the race. Incidentally, the Grand Challenge is not a ‘race,’ though that is the word that comes closest to describing it. Its not a race because it is against the law to race on (some of) the OHV areas. So the Grand Challenge is a Challenge. The name Slash X Cafe is a bit curious. The middle letter has interesting connotation. If there was any such conception in the teams mind it was put to rest by a dirt-biker who visited us. I cannot recollect the exact words he said. The overall implication was that the waitresses in the cafe can put a New York bouncer to shame. Most of them have few front teeth missing, and you better not ask any of them why. The Slash X Cafe is also in the middle of nowhere. It is the only building, if you can call it so, in a valley for as far as one can see. Once you get there, you better be prepared for their asking price for food and drinks. My teammates have already gone to Slash X. I am in no rush to go ‘nowhere.’ Besides, I do not know if I’d find connectivity there. Before I head there I want to make sure that CajunBot fans get their daily dose.

Arun Lakhotia

Team CajunBot’s Sim Team

Friday, March 12th, 2004

The Sim Team, as the trio Nitin Jyoti, Suresh Golconda, and Arun Pratap, is called is one of the reasons for the remarkable performance of Team CajunBot. The team was put together in early February, triggered by a comment made by Tony Maida. Tony mentioned that most robot development kits come with a simulation environment in which one can first test the software before it is loaded on a robot. That made perfect sense. The team started thinking about how to build a simulation environment that can help in debugging our software. A simulation environment, for those not in the know, is a computer program that (in our case) will ‘act like a real world.’ It will be a virtual world, a world that solely exists in machine, but it would generate the same stimulus as a real world. Computer games, like those for Nintendo, X-Box, etc., are examples of virtual worlds. Arun, Suresh, and Nitin were brought into the team in early February to develop a simulation environment. By the time we arrived in Hesperia, CA, the simulator environment was ready. This environment (a program) had a virtual 6-wheel car with all the sensors that our real robot had. With a little bit of work we were able to connect our robot software with the simulated world. Now we could test our robot in the comfort of our hotel room, instead of running around in the desert. On March 8th when we had to quarantine our robot, we reverted to the simulator. Our first integrated system, ready on March 9th, was tested on the simulator. The testing revealed several errors, that we corrected. When our bot made is maiden run on March 10, we had a very high degree of confidence that we’d fare well. Though we had not had the chance to run the bot (because it was quarantned), we had run the bot in our simulator. It is because of the simulator that we could have a near-perfect performance in just three test runs.

Arun Lakhotia

CajunBot qualifies to be at the start line

Friday, March 12th, 2004

CajunBot has qualified to be at the start line in Barstow. She is ranked 7th in the list of 15 selected. CajunBot fans: rejoice and paint the town red. It was a defining moment of joy, accomplishment and glory, as the media, competing teams and fans rejoiced in a roar as CajunBot was announced as the 7th of 15 teams to qualify to line up at the starting line. It’s been a tedious endurance race simply to get to the start line, with many members of the team staying up all night last night and with little sleep prior. For the moment, satisfaction is at hand.

Arun Lakhotia & Mike Spears

CajunBot does two-steps, and then snails through the track

Friday, March 12th, 2004

Yesterday, March 11, was a gut wrenching day. While we await a decision by DARPA in about 2 hours, I’ll share with our fans our moments of excitement and anxiety. Late March 10 we decided to do the unthinkable. Make substantial changes in the underlying operating system. Our analysis indicated that in order to be a successful challenger, we must have those capability (NTP protocol to time sync multiple machines and devices). It was a calculated gamble. We figured we did not have much to lose if it did not payoff, but a lot to gain if it did. The gamble did not pay off. So we had to revert to another plan. We developed a new strategy for avoiding obstacles without slowing down. Our base computing hardware is a 2.0 Ghz AMD processor, a capability one can find in a garden-variety home computer. With this minimal computing power, we really have to be very creative. We went into QID Round II with the modified strategy. CajunBot loved this strategy, for we could see it doing Louisiana two step around the course. In its own merry way, it meandered around as though unaware of any bumps, walls, or other obstacles. For those who are not familiar with Louisiana two step, its a beautiful and uplifting dance. Typically a large group of couple dances around the dance floor weaving around, often turning 360 degrees, sometimes bumping into other couples. CajunBot did several 360 degree turns, and as it approached the parked van, it figured it seem to have decided to ask it to dance. That turned out to be a poor timing for dancing, for DARPA does not permit two vehicles going around the track. You can dance around the track, the policy goes, but you must dance alone. It was around 1pm when this happened. With this performance, we were clearly out of the race. However, the final whistle had not yet been blown. It was time to put on the Ragin Cajun hat and go for it again. Within an hour we were back on the track, this time our strategy was to revert CajunBot to a slow and thoughtful pace, just like the other vehicles. Its a shame, but social norm (DARPA rules) and peer pressure does come in the way of fun. Our QID round III was around 2:30pm. CajunBot could sense it was the final round too. This was not the time for charming the crowd or the van parked on the route. The bot proceeded at a snail’s pace, around 2.5 miles per hour, the kind of pace all other bots have maintained. It went around the track taking each step carefully and thoughtfully. As it went around we could notice it had an excessive veer towards the right. Since the track is anti-clockwise, that also means it cames too close to walls for comfort. Nonetheless, it always corrected itself at the last minute. CajunBot went through gates, past van, past cars, through the arches, through cattle guard, went around turns, and just before the last turn came very close to the cement wall around the course. It was barely 1/2 mile away from the finish line, when it stopped in front of the cement block. This section of the track makes a nice curve, the likes one sees on NASCAR races (and this is a NASCAR race course, so its not a coincidence either). As it stopped and contemplated near the cement wall, it raised the curiosity of the DARPA Chase car official. The waiting went on for a minute, or at least that’s what they fealt. The official decided to put CajunBot on ‘pause,’ and stepped out of its truck to check out what was going on. The bot was very clear from the wall, there was nothing blocking it. As the official checked around, he sensed the bot lurch, attempting to move. The official got on the radio yelling ‘Bot out of control’, ‘Bot out of control’, and ran to his truck parked about 100 feet away. All the radios were buzzing. There was a lot of commotion. The official yelled on his radio ‘diabling bot,’ ‘disabling bot’. Within a matter of seconds CajunBot was taken out of its thoughtful travel and put to a stop. I can sense CajunBot thinking “Its hard to live in an adult (DARPA) supervised world.” You’re penalized for being reckless and fast, you’re penalized for charming and frolicking, and you’re penalized for being thoughtful and slow. In the next hour we will have DARPA’s decision. So far only six teams have completed the whole course. Five teams did not show up. Three had difficulties in starting the course. That leaves a very narrow field of players. We are proud of what we have achieved in a matter of three months. Our final run was close to the first run of the lead players with years of experience, and months of testing. We have closed in very fast.

Arun Lakhotia

CajunBot Fans, Today is Your Day!

Friday, March 12th, 2004

Good Evening! Here it is Thursday – nighttime again and qualifying time is up….Trouble has this Excursion running about 90 and we’re leaving the valley and headed for the high desert. Well, let me tell you what happened today. After leaving the California Speedway last night in great spirits and getting a good meal and decent night’s sleep, today was tough – real tough. On the first attempt, CajunBot’s wits were still sleeping in the garage. Team CajunBot then rescheduled another try around noon. By the time the flag fell, CajunBot sped right out the gate on down the track. After analyzing the field by viewing left and right, The Bot centered itself on its GPS coordinates and rolled down the course right on axis, until alas!, right into the back of the minivan that’s situated on the route. No damage, but DARPA e-stopped us. Back to the garage to sort it out. Now I can’t give you the actual explanation of the difficulty, but let’s just say that the Team was stumped for quite awhile. The final try came at about 5:00 pm. Same start, but perfect maneuvers and the most distance than CajunBot had gone all week! All the way to the far end of the track, where the Bot was paused as it got to close to the wall. It moved a bit further, and then waited for the Team to collect it. By the time the Team returned with the CajunBot, aDam and I had two steaming pots of jambalaya ready for them. Now, with Clotile (the 6 ft. crawfish) and the fine red UL school flag adorning the Med Express chase team trailer, all of the competition got a chance to see a real Louisiana tailgate party/team meeting. And quite a few guests enjoyed a taste of Cajunland as well. What started out as a really tough day finished just like at home – supper! Well, it’s over now and tomorrow morning (Friday) at 8:00 am, Team CajunBot will know if Louisiana will be guests in Las Vegas or if the little “CajunBot that could” will compete in the Grand Challenge for the Grand prize – one million US dollars! I’m sure that someone will update this journal before noon CST. Whatever the outcome, Team CajunBot has carried its fans into unchartered territory in the field of Autonomous Robotics Technology. So today, enjoy this moment – it yours!

Team CajunBot
Danny J. Majors, djmajors@aol.com

Team CajunBot rebuilding Linux Kernel hours before QID round 2

Thursday, March 11th, 2004

Today, Thursday, we are scheduled for Round II of QID. Our analysis from data of Round I indicated that our bot may have been stopped because of two reasons (1) unsynchronized sensors and (2) too much disk writes. Patrick Landry, the system administrator, on the team ruled out reason number 2. He concluded that it was not the amount of disk writes, but a lose connector on disks that was the cause of the problem. Cause 1 is still critical though. If our sensor data are not synchronized, then we cannot relate data from multiple sensors. To give an example in human terms, if what the eyes see cannot be time-synched to what the hands feel, then we’d never be able to drink coffee. We have known the issue. It was not a priority before because we really did not have all the pieces together. The issue has become top priority now. Yesterday evening we decided to bite the bullet and rebuild the Linux kernel to introduce the necessary capability. The decision has its own risks. It is not easy to integrate all the hardware drivers (for joystick and other sensors). As one would expect, we are now running into the problems. As I write this, Patrick is sitting outside the garage on the asphalt, trying to find the reason why the laser range finder (from SICK) is not working. We have spent enormous amount of time and money on these laser range finders. Every time we do any changes, they seem to break. But it appears we have no choice. All other teams are using the same product. Coming back to the building kernel, Team CajunBot has been running against time from Day 1. So it is no surprise that we will take the risk a few hours before QID Round II. Besides, we need the new capability to improve our system. And there is no real point in doing a test run with a system without adding new capabilities that improve upon previous run.

Arun Lakhotia

A Moment of Making History….

Thursday, March 11th, 2004

Good Morning! I must say that it is just awesome to be the center of so much media attention. There is a sense of accomplishment to be surrounded by the United States military, defense contractors, to recieve the attention of the most well-known scholastical institutions in the world….,but my proudest emotional feeling comes from being a part of Team CajunBot from the University of Louisiana. The announcer has likened the invention passion of these students, and this resulting Grand Challenge to the Wright Brothers work, a tiny part of history in the making. I am sense some arrogance (probably well-deserved) among some of the entries, but the humble determination of this Team has attracted worldwide media, the complements of DARPA, and the respect from most everyone else here. Louisiana goodwill will always produces the real winners! GEAUX CAJUNS!

Chase Team
Danny J. Majors, djmajors@aol.com