October 30th, 2012
Here is a quick recap of the two previous pool tests that were recently carried out:
The first pool test did not go quite like we wanted it to. We had some networking troubles with the computer, so we went back to the lab to correct this problem.
The second pool test went exponentially smoother than the first. One of the only problems that we had with the second pool test was that we could not get the forward camera to work, so instead, we moved the bottom camera to the forward position. When we put the robot into the pool, we ran gate mission. After tuning the vision algorithm, the mission worked very well. We wanted to test the path but tested with the buoys, because we had already moved the bottom camera to point forward. While working with the buoys, we found one problem with the buoy vision entity, but hopefully we will fix this problem in preparation for the third pool test coming up soon!
Stay tuned for more URC news to come!
And remember to follow us on Twitter and Facebook !
October 14th, 2012
This may be the biggest year yet for the NCSU Underwater Robotics Club’s mechanical team. With a total of thirteen people working on the mechanical projects this year, it’s a good thing we came back with ambitious goals for the mechanical design this year. We were inspired at competition to once again start from scratch with our mechanical design. Seawolf VI will feature a modular chassis that will hold the cylindrical hull, both of which will be made of aluminum. The modular design will allow us to switch out different plates to allow mounting for various mechanical systems such as camera enclosures, thrusters, torpedoes, or the grabber. The design for the chassis and hull has been completed and machining the aluminum should start within the next month. Seawolf VI is currently on schedule to have its first pool test in the beginning of 2013.
Remember to follow us on Twitter and Facebook !
September 27th, 2012
It’s the new school year, which means another year of Underwater Robotics! Over the summer, the club went to San Diego to test our underwater abilities in the annual Robosub Competition. This year, we are proud to hold 7th place out of 29 other teams, which is the highest we have ever ranked in the history of URC! This is exciting news and gives us bigger aspirations for the 2012-2013 year. A quick congratulations to Cornell University for placing 1st in the Competition!
So far, in preparation for the new year, we have recruited over 15 new members, so our club has grown exponentially. We have many goals set before us, and we have organized small groups within the club to pursue these goals. Details are to come with later posts, but for now, here is what we have been working on so far:
- Electrical: This year’s electrical system will be a new revision of last year’s system. These revisions will allow the electrical board to support another thruster. Also, minor changes will be added such as voltage and current monitoring, and depth readings.
- Acoustics: This year we received the NCSU Undergraduate Research Grant, to pursue underwater acoustics. The acoustics system will be a modified version of the incomplete system from the previous year. Our goal this year will be to locate the pinger using multi-lateration, which is similar to GPS triangulation software.
- Software: As was last year, we will be working extensively with the simulator, but with Seawolf V still in working condition, we will have plenty of time to debug software during pool tests. With our software team growing, we expect to have more comprehensive mission and vision software. In addition, the diagnostics display will have live-feeds from Seawolf V, as well as debugging streams through our video routing server.
- Mechanical: Our goal this competition season is to have an aesthetic, as well as functional, mechanical design. Seawolf VI is currently under construction. Our plan is to have machined aluminum parts for the robot in its entirety.
Stay tuned for more as we continue building our robot.
Remember to follow us on Twitter and Facebook !
October 18th, 2011
The RoboSub Competition has finished with NCSU making its best finish ever, placing 8th of 27 teams and making it into the finals for the first time. Congratulations to SONIA for getting in first place with an amazing final run!
We’re already excited for next year and have begun work on Seawolf V. Most of the robot will be redesigned and rethought completely. We’ll post plenty of details in future blog posts, but for now, here is an overview of our progress:
- Software – We have already re-architected the vision system since competition. We now have the capability to process binocular vision data in two different processes. Thanks to our recent Intel sponsorship, we have a brand new i7 processor with 4 cores and 8 threads! This is a huge improvement over our old single core Atom. Thanks Intel!
- Electrical – We have started peripheral communication, power distribution, thruster control and all other electrical design entirely from scratch this year. We are throwing away all our Arduino microcontrollers and instead using the MSP430.
- Mechanical – We have already modeled most of Seawolf V. Thanks a lot to TechShop, who is letting us use their facilities to construct our new robot. Thanks to them, we will be machining our own ABS plastic for the frame. If you haven’t heard already, the newly designed Seawolf V will look like a tie bomber!
- Acoustics – We have two different acoustics projects in progress this year. One is an undergraduate research project by Chris Thunes, which will utilize a Blackfin DSP and an MSP430. He already has his board’s first revision in the debugging stage. The other acoustics project is going to utilize a Gumstix after doing filtering in analog.
Look out for more blog posts in the near future about Seawolf V as we continue development. We’re going to be keeping busy so remember to follow us on Twitter and Facebook!
July 17th, 2011
Alright folks, we just completed our final competition run. Seawolf charged the buoys after passing cleanly through the starting gate, followed up with a stylish spin to align with the hedge path, cruised through the hedge, and perfectly targeted the love letters. I could say our veins are still pumping with pure adrenaline in hopes our perfect finals run isn’t quashed by another team’s run.
In a chaotic world, however, you have to expect curve balls. That’s exactly what we were handed. So what exactly happened? Before the final runs began, the course was changed significantly- Most notably, the positions and heights of the buoys were changed enough that the positions exceeded the hard-coded “limits” set in our code. Seawolf clearly saw them, but dismissed them as false-positives. This was a painful blow, and really hits home for everyone. Some of us have been through this process before, and know to expect and have to be ready to eat the worst. But for the newer guys, it downright sucks in every way. But that is what keeps us on our feet- We have once again brought the team closer to victory than any year in the club’s history. If we gain too much ground, what will be our motivator for the next year? What keeps us from getting too comfortable? Too complacent?
The team is still in high spirits, and after our less than stellar run is off everyone’s minds, we’ll realize that where we are this week, ranked in the finals, far exceeds where we ever thought we could be several years ago. Everyone single person on this team needs to pat themselves on their backs, and lets not forget our advisors and faculty working voodoo magic behind the curtains- What a hell of a team.
July 17th, 2011
These are older photos (from Friday), but we figured they’re post-worthy!
We’re about a half an hour out from the start of the final rounds, and the late-summer morning gloom is just beginning to burn off. The software guys are adding a final high-gloss finish to their code, while there is talk about dropping the robot in the small dolphin pool to wake Seawolf up with a last-minute hardware test.
Keep your fingers crossed, and keep your eyes on the final LIVE webcast over at robosub.org!
July 17th, 2011
Seawolf IV during practice time
For the first time ever, NCSU Seawolf is in the FINALS!! With a LOT of hard work on vision tracking, calibrating, debugging, and begging seawolf to do what we want, Seawolf IV had a phenomene ranked 7th out of the 28 teams. The top 8 teams made it into the finals this year. This is the first year that our team has made it into finals and we are ecstatic. The final rounds will be webcast (link below). We have been tweeting a lot (even while on the dock when the robot is in the water). We would like to thank you all for your support and we look forward to some good competition in the finals tomorrow.
Finals begins being webcasted at 1PM PDT(4PM EST). We will be competing directly after Cornell University.
Robosub website (webcast can be found here): http://robosub.org/
Twitter (will be updating throughout the finals):
July 16th, 2011
Seawolf with the Army bot
Today is the second day of actual competition qualifications. Tonight, the judges decide who moves on to finals day tomorrow. We have just had possibly the best run of NCSU robotics history. Seawolf IV swam through the starting gate, aligned with the path, hit a buoy, aligned with the next path, and went through lovers lane. It then started the bin dropping mission by finding the double path. Seawolf had aligned with the next path but it was the one headed towards the torpedo launch area instead of the bins.
We have had a few things go wrong, but we have so far been able to fix almost everything. Our Ethernet port gave out on Seawolf’s netbook but were able to find a local store (Fry’s) that had USB to Ethernet and it is now working properly.
We are hopeful about getting into finals but either way we have done extremely well. Seawolf is getting into the theme of the competition (roboLOVE) and is now in a relationship with the Army robot.
July 15th, 2011
We did well today during our pool session. We managed to go through the gate, align with a path, and hit the red buoy. Our robot could not see the next path and drove forward. We accomplished this on our second attempt (the first attempt: Seawolf went through the gate and didn’t see the first path). After locking in our score, we did several more autonomous runs so that we’d have more video footage to analyze.
This morning we also did our static presentation. We could definitely have put more time into our preparation and use some practice on our presentation skills. That being said, everyone had the opportunity to speak about the robot, which is a good learning opportunity for new members. Unfortunately we scored 16th in static judging (it’s a coincidence that in the image here there’s a 16 next to our name…the numbers on the left do not correspond to static judging).
In the late afternoon we were given another opportunity for a practice session. Prior to the practice session we began having connectivity issues over our tether. We determined that it was a computer issue and not a connector issue. Since connectivity was intermittent over the onboard ethernet jack, we decided to purchase a USB-to-Ethernet dongle locally. The dongle arrived on site just before our practice session (10 minutes to spare). We weren’t able to get drivers working with linux in such a short time frame, and luckily after a few reboots of our computer the existing connection began working again. We’re hoping this won’t be an issue tomorrow (Saturday), and we’re hoping to also have the USB backup solution as an option.
Keep up with the team on Twitter: @NCSURobotics
July 13th, 2011
San Diego 2011
Yesterday the NCSU team arrived along with 27 other teams in San Diego, CA for the annual RoboSub competition. We picked up our robot from the shipping hub and made it just in time to our 2pm orientation meeting. Today marked the first day we were on site the TRANSDEC facility at SPAWAR Systems Center Pacific on Point Loma.
We had two runs today. During the first run we were plagued by some FireWire camera issues. Our plan for our first 30 min practice run had been to collect some video footage to work with. Unfortunately we only got a few frames from our forward facing camera until we had issues. We did collect some other images from the “bins” on the bottom of the pool with our down-facing camera.
Our second pool test was in the afternoon, and by that time it seemed like our camera issues were worked out. We touched up the underwater connector a bit in addition to trying a few other things. We still aren’t quite sure why the connection keeps going out, but perhaps is an X-forwarding issue. During our second test time, we collected the footage we needed, and did two autonomous runs. We did one with the tether disconnected such that the judges were satisfied that we pre-qualified. To pre-qualify for Friday, teams must have their robot autonomously navigate through the gate while being fully submerged.
We refueled for the evening at Vons (the local grocery store) and Subway. We’re getting some computer vision pool testing time in the hotel pool this evening. Our Thursday goal is to successfully accomplish the gate, path, buoys, another path, and perhaps get to the dropper bins. Things are going well in San Diego this year! We definitely don’t miss the humid N.C. weather!