Archive for the ‘Uncategorized’ Category

Communication Change

July 23, 2009 - 10:26 pm 2 Comments

It has been over 3 weeks since matt announced the communication change to use WANEM. He has yet to post the detail settings and configuration that we would use to test our robot. Since we had already setup our link delay and bandwidth restriction in a manner similar to the WANEM, we decided to share with you guys what we feel is the best way to configure it.

comm summary

(more…)

July 10, 2009 Beach Test

July 11, 2009 - 5:25 pm 2 Comments

Yesterday, we completed a 2 hour beach test to check several new features on the vehicle. We added a fourth camera, significantly increased the power of the shaker and added sensors to check the level of regoligh in our storage bin. The camera was OK but not a absolute must have. The new shaker helped tremendously in redistibuting the sand in our bin. The sensors had some issues with transient signals.

Driving skill has  increased and dump to dump cycle times have been coming down steadily. In our “run for record” for the day, two remote operators running on camera information only managed to complete a circuit in 4 minutes and 30 seconds. The run collected ~64 kg of sand which translates to 76 kg of regolith at the same volume.  Not too bad!

Of course, we will probably not have the luxury of an unobstructed path during the competition so some time will have to be allocated to avoiding rocks. Still, the actual cutting time of the run was only 1 minute and 50 seconds so there is room for improvement. We are only using 12% of our cutter power capacity so we hope to have the same material removal rate in the compacted regolith.

A new video has been uploaded to youtube of the 4.5 minute run.

beach day

June 25, 2009 - 1:34 pm 1 Comment

We are headed to the beach for a day of testing tomorrow 6/26/2009.  We will be meeting up at the shop around 1pm, and plan to be at Dockweiler State beach (google map link) around 2pm. We should be there for several hours, so come and join the fun! We will be in the payed parking area. However, you can park on the street for free and walk down.

Several new features have been implemented, and will be tested on this day:

  • New Control Panel GUI – many new useful widgets including intelligent video presentation, current consumption plots, and vehicle mechanical state displays.
  • Multiple Controllers – Allows for multiple computers to concurrently control and observe vehicle behavior.
  • Cameras on Servos – the front cameras are now on servos and can be remotely actuated.
  • New Rear Camera Position.
  • Realistic Moon Communication – simulated time of flight delays to the moon and 1mbit bandwidth limitation.
  • Competition Test Area – actual size test arena and collection box.

Preparing for the 2009 Regolith Excavaction Challenge

June 17, 2009 - 10:31 am Comments Off

We are busy at work preparing for the 2009 competition. We have made several design changes to account for the new rules. Although we feel we had a very good entry in last years competition (with the exception of the flawed h-bridge chips), we feel that this year we will be even better.

We have designed new motor controllers using much more robust and modern chips. Our original design was very modular, and once a new pcb was created it simply plugged in and worked. If only we had a time machine. The new controllers work flawlessly, and performs exceptionally well when under stress.

The big rule change this year is the removal of the autonomy requirement. We had a very advance and well implemented system for automation. However, we feel that we can be much more competitive with a human in the loop making more decisions that software, even with the 4 second round trip delay. As such, we have removed the ladar, and replaced it with 3 cameras.

The removal of the ramp in the competition also caused a slight design change in the vehicles collection box. We feel driving up and down the ramp is inefficient and a huge waste of time. Our new design allows for a quick and easy transfer of regolith from the vehicle to the collection box.

The last major change has was the wheel size. Based on performance testing and with weight considerations, we decided to lower the radius of the wheels. We are able to move much faster, and hope to exceed the minimum requirements for excavated mass.

2008 competition results

June 17, 2009 - 10:12 am Comments Off

We were unable to make a comeback from our major hardware failure. The end result was that we mistakenly choose an h-bridge chip design that was very old, and very sensitive. some note from a post-mortem done be a team member:

The undershoot voltage capability for the chip is significantly below what today’s chip can stand. Of particular importance is the -0.3 V limit on both the logic input pins and the low side gate driver pin. Most modern chips can take -2.0 V at these pins. It is one of the reasons they are discontinuing the chip.
The traces showed that the logic inputs were reaching -0.5 V for ~20 nsec. That is too fast for just about any diode to help with but he thought that the short period of time would make the transients a non-issue.
The traces also showed that the low side gate driver was hitting -0.7 Volts. That would be enough to slowly damage the chip until it latched up. If one low side latched up and you were controlling the other side, the chip would simply overheat. If you were controlling the side that was latched, it would short.

When we changed motors, we changed the inductive load and the current transients. When we run both motors simultaneously, the probability of latch up increases.

Could it have been fixed? Maybe. The addition of some ultra-fast diodes on the low side gates may have helped. Ultimately, we needed to play with the trace routing to minimize the coupling between the power traces and the gate traces. Without sophisticated software, it would have been trial and error. It would have never been an issue with a modern chip.

We did bring the vehicle to the competition to show off all of our hard work. People seemed to be very impressed with our design and simulation. We will definitely be there for next years competition!

t minus 1 week

July 27, 2008 - 2:11 am Comments Off

Things have been crazy in the past week for us. Our day jobs are out of sight and mind as we focus on being ready for the competition.

In the last week we have made the following accomplishments:

  • vehicle takes into account mining history
  • intelligent mining loop locations based upon blocked grid
  • ramp mode (go to ramp after mining fixed distance, go up ramp backwards, dump, return)
  • removed ir sensors (real life sensors are too noisy)
  • fault management for several subsystems: ladar, i2c, gimbal, pathplan
  • started PHD fault management (ladar intensity)
  • added more bump sensors
  • gimbal leveling (the vehicle autonomously drove up and down the ramp while maintaining PHD lock)

 

Final assembly is going well. We are very close to the weight limit. As of now, the BFD measurements are:

  • 4 x wheels w/drive @ 17.0lbs each (68.0lbs)
  • Frame including electronics (31.6lbs)
  • Collection Box including chain and buckets (15.2lbs)
  • Collection Box drive (26.7lbs).

 

That’s a grand total of 141.5lbs. We are very close to the limit, and that does not include our marker weight.

power board

July 26, 2008 - 12:52 am Comments Off

the powerboard has been mounted and wired up. thanks to jim for making it look so pretty and clean ;]

photo

colour!

July 25, 2008 - 4:27 pm Comments Off

the new frame has been painted… look how pretty!

photo

photo

smoke and fire

July 23, 2008 - 9:39 pm Comments Off

last night while doing some performance tests with the new ladar we had a small mishap. the front right motor controller cratered it’s bridge chip. not quite sure why it happened but the power connector might have come loose. we swapped in a spare and the vehicle is fine now. jim repaired the board and placed it into the pool of spares.

photo

new frame with cutter

July 23, 2008 - 9:55 am Comments Off

the new lighter frame has been completed and the cutter integrated. migration of other units into the new frame is coming soon.

photo