Congratulations to UF’s NaviGator AMS team, winner of the 2016 Maritime RobotX Challenge (www.RobotX.org) in Hawaii. See www.NaviGatorUF.org and www.twitter.com/NaviGatorUF for more info.
UF’s Team NaviGator AMS, winner of the 2016 Maritime RobotX Challenge in Oahu, Hawaii
With it raining cats and dogs, PropaGator continued its testing in the Florida pool. Testing today mainly consisted of navigating a marked channel. However, after two hours PropaGator lost one of it’s motorcontrollers followed shortly by the boat’s wireless router. Bringing the boat ashore, we discovered that both pontoons had some water inside. The rain managed to get through a broken portion of one of the reactors and through a partially closed lid on the boat. Returning to th
We continued our navigation of buoys until our LIDAR mount motor gave out. Ensuring that we used the most of our pool time, we switch to GPS navigation testing. By the end of the day PropaGator was able to turn towards a GPS given point and drive until the boat was on top of the point. We also found out that 40A fuses were not enough for our boat. 50A fuses have now been installed.
Today, the weather in Florida was beautiful and we had little cloud coverage. The boat performed well and we were able to get in a solid four hours of pool testing. By the end of the day, the boat was able to identify pairs of green and red buoys. After identifying the pair, a desired location was placed in between the two buoys. We were able to visualize the desired point in RVIZ but had problems getting the boat to travel to the location. PropaGator had an unhealthy attraction to the red
Today, at the UF pool, PropaGator attempted to navigate through a channel marked by red and green buoys. Due to a software change, the boat was immobile and could not be fixed at the pool. In an effort to not waste pool time, the boat was pushed through the channel and the video and LIDAR data recorded. The data will be used in the lab to further refine the channel navigation software.
Today we measured PropaGator’s thrust, simulated the speed channel, and bumped more buoys. PropaGator did well on the thrust test and proved it could maintain a straight trajectory through the buoy channel. The boat can now recognize and bump blue, red, green, and yellow buoys. Next step, beginning work on maneuvering through the marked channel.
Detecting objects with a camera or a LIDAR is one one thing, using both electronics as one sensor is another. This weekend, PropaGator successfully fused both the camera and LIDAR as one sensor used to detect buoys in the pool. PropaGator can identify objects with the camera and acquire the object’s bearing and range, relative to the boat, using the LIDAR. Overcoming this hurdle was a big challenge for the team. Teaching PropaGator to maneuver through the channel will be downhill from
Since we moved the camera and LIDAR, we had to re-calibrate both sensors. We have no plans of moving them again and should not have to do the checkboard calibration again. Saturday’s weather cooperated with us. The sun was out most of the time and we were able to gather some very good LIDAR and video data. Out goal today was to identify a buoy and hit with the boat. If we can hit a buoy then we can dodge a buoy.
Before today, we used an extended Kalman filter to determine our location and orientation. The extended Kalman worked decently but PropaGator would sometimes lose its position and drift wildly. In an effort to get a more accurate position, we tried using an unscented Kalman filter. The boat had a few issues with the filter so we logged the data and digest it in the lab. After testing the filter, we moved on to detecting the blue buoy. We used both the LADAR and the camera.
Today we continued to test our controller and monitor our Kalman filter. PropaGator made several erratic maneuvers during station holding testing. We were able to log the inertial navigation system data for analysis at the lab. Hopefully we can figure out what happened… Next, we played with our buoy detection using our LIDAR and camera. During our testing, our tilting motor failed. From the moment we turned on the motor, the audible sound of gears rattling could be heard. The noise c