Country: Denmark
Year Submitted: 2018 (Updated 2019)
University: Aarhus University, School of Engineering
List of Team Members (with year of graduation):
Alexander Fürsterling, 2021
Simon Sejer Pedersen, 2019
Mathias Skovby, 2021
Lasse Vesterled, 2019
Faculty Advisers:
Claus Melvad
Main Contact Email Address:
Mathiasskovby@gmail.com
Description of ARCAB:
Expedition Greenland 2019:
Title: ARCAB - Arctic Research Centre Autonomous Boat
Description: ARCAB is an autonomous boat to be used in polar waters to obtain data like depth, current, salinity and temperature measurements. This data is used to update climate models, especially with focus on the impact icebergs have on global climate change.
Products:
Software |
NI LabVIEW 2016 NI LabVIEW 2016 MathScript RT Module NI LabVIEW Real-Time Module NI-VISA |
Hardware | myRIO |
Other |
2x Pulse 22,2V 6000 mAh 6S LiPo batteries 2x decouplers for batteries in series 2x Graupner GM Ultra Race brushless motors 2x Reely Sky-Series ESC 2x clutch, stern tube and propeller 1x DC-DC converter 19V 1x DC-DC converter 12V 1x MAX232E microchip - RS232 to UART 1x DualNav150 GPS 1x modCMPS compass module 1x Scanse Sweep Lidar 1x FrSky Taranis x9d Remote controller 1x FrSKY X8R receiver 1x WiFi router 1x Blinking light, red
Miscellaneous: Glass fiber and epoxy Gaskets Wires, capacitors and fuses Laser cut brackets and hatches |
The Challenge:
The amount of freshwater from melting ice led into the oceans in the Arctic area has a big impact on the global climate. Small changes in the amount of freshwater influences the models for the world climate considerably, but the estimates for the melting ice are not very accurate today. Measurements of the currents, depths and salinity of the waters in the Arctic can help improve these estimates. Such measurements are interesting to make in lakes, fjord and around melting icebergs.
Depth measurements can be used to determinate what kind of vegetation is growing in a certain location. The local flora and fauna around Greenland, influence the local and global ecosystem. Information about water currents and depths can also be used to estimate when iceberg are stranding and how fast they will melt.
Daniel Frazior Carlson (called Dan Carlson) from Artic Research Center is using the RiverSurveyor M9 to measure depth and water current in the oceans and fjords around Greenland. To do this he rents a boat and crew to sail around with this measuring equipment. From the side of the boat he can hold the M9 down in the water with a pole. This solution is not efficient in various ways. The boat and crew is expensive, and other research colleagues would also like to use this boat. The boat is big and therefore difficult to maneuver in small waters, and its is hard to maintain a constant slow speed. This limits the quality and resolution of the depth and water current measurements, which can be mapped. Furthermore a manned boat may not come closer than 2 km to a glacier due to safety concerns. This challenge is the background for the problem this paper concerns.
The Solution:
ARCAB is catamaran where the RiverSurveyor M9 can be installed onto. RiverSurveyor M9 requires a battery pack and GPS antenna to be carried alongsige, why these too are installed on ARCAB. ARCAB can be either controlled by an RC-controller or it can be in autonomous mode. In RC-mode, ARCAB, behaves like a normal RC boat, where the user can control speed and steering. In autonomous mode, ARCAB requires a set of GPS coordinates to be used as a route. These coordinates are planned in the router planner for drones, MissionPlanner. ARCAB has a GPS antenna, from which it knows position, direction and speed over ground. These informations and the route specified are used to control direction and speed of ARCAB. It will follow the route with however many GPS-coordinates specified and stay at the last coordinate when the route is finished. The route can be updated mid-mission. If ARCAB is sailing towards an obstacle, the lidar will detect this, a change in direction is calculated and ARCAB will sail around the obstacle.
ARCAB's catamaran design is designed to have a high draft and a wave piercing stern. This reduces the impacts of wind and waves. The deck(link between hulls) is made from PVC-core material with glass fibre on each side. Aluminum finns are installed to protect the propellers in water from seagrass and debris and makes it possible to place ARCAB on land without damaging the propellers. These doubles as cooling solution. At the back of the deck, a waterproof junction box is placed where the myRIO and other equipment is installed.
Refer to upper figure for following list:
1. Extra mounting possibility for salinity and temperature measurement equipment. Included to ensure ARCAB is suited to multiple types of missions rather than just with the RiverSurveyor M9.
2. Hatches for waterproofing hulls, which are accessible. In the hulls, motors, ESC, decouplers and batteries are placed. See to next picture
3. Handles for easy handling of ARCAB for the user.
4. Mountings for GPS antennas. The larger one is for RiverSurveyor M9, while ARCAB is using data from the smaller one.
5. The RiverSurveyor M9 mounted so the bottom part, the data acquiring, is submerged 10 cm into water.
6. The junction box. See related figure. In front of the junction box, whilst not visible on upper figure, the M9 battery pack is installed.
In front of the RiverSurveyor M9, the lidar is placed inside a transparent tube, offering water proofing. On top of the lidar a light is installed, making ARCAB easy to spot even in daylight.
Explanation to junction box:
1. Connection tube from left hull
2. Connection tube from right hull
3. myRIO
4. Spot Trace, emergency position transmitter
5. DC-DC converter 12V
6. Wires from left hull
7. Wires from GPS antenna
8. Wires from right hull
9. RC receiver
10. RC receiver antennas
11. Power for myRIO 12V
12. USB connection to router
13. 5 GHz wifi router antennas
14. DC-DC converter for router 19V
15. Circuit board
Software
ARCAB is controlled via the following code:
Explain the benefits using LabVIEW and NI tools.
LabVIEW offers a high level programming language with intuitive setup and a wide variety of capable configurable devices, notably for this project, the myRIO. Using LabVIEW, even developers with sparse programming experience can make working prototypes(ex. ARCAB) and helpful applications.
Level of completion (beta, alpha, or fully functional)
ARCAB is a fully functional prototype. At time writing it is being transported to Greenland, where it is to be used for several missions from end-July 2018. ARCAB will be used in both RC and autonomous mode.
Time to build
Project spanned from start february to mid-may, equating approximately 3.5 months, however most work was done during the last 2 months. The workload has been around 200 hours per group member, excluding reporting and this submission.
Additional revisions that could be made
1. Enlarge ARCAB. ARCAB is heavier than initially designed. Making the hulls larger will offer more buoyancy and thereby more lifting capacity. It should not exceed the dimensions of 1200x900 mm.
2. Actuators for propulsion are over-dimensioned due 6S LiPo batteries being used (requirement in our case). Each could be downsized to delivering 300 W without problem.
3. Wireless communication has a range of approximately 200 m. It would be nice to have 2 km.
3.1 Install alternative, more simple and beefier WiFi router to communicate at 5 GHz.
3.2 Install 433 MHz transceivers in both ARCAB and computer for communication. The boost and sensitivity should be above 120 dBm combined.
4. When RC communication is lost, ARCAB will keep doing the latest information it got. A way to solve this is to use the digital pins on the receiver to communicate whether there is signal.
5. The code is quite heavy in terms of time consumption, why it would be smart to work with optimisation. A part of this could be to make use of the FPGA. A direct plus by this would be, that the lidar could be allowed to make more samples, making the obstacle avoidance more bulletproof.
6. The lidar is placed inside transparent tube, however spray from the sea can make it blind. Possible solutions are to install a shield, minimising spray and using a water repellent wax.
ARCAB featured in an expedition to Greenland in January 2019 where scientist Dan Carlson from Aarhus University, Arctic Research Centre, used ARCABs capabilities to map several fjords and bays around Nuuk, Greenland. During this expedition, ARCAB was deployed from both vessel and shore. It was used in both RC and autonomous mode. During the expedition, only minor adjustments was conducted on ARCAB, relating to direction calculations from GPS signals, even though the environment in which it was deployed was both colder and windier than described in specification of requirements.
ARCAB was deployed near Nuuk, Greenland's capital, in early January 2019. The 1st deployment took place 7 January 2019 in Kobbefjord in a shallow (2-7 m) embayment. ARCAB carried out two autonomous missions, mapping bathymetry within meters of the shoreline in depths < 2 m. The second deployment took place 8 January 2019 in a small bay in southern Nuuk. ARCAB operated autonomously and under remote control in air temperatures of - 15°C to -10°C and with steep, choppy waves. ARCAB was deployed from a small boat and from shore and acquired bathymetry data close to navigational hazards that would make the operation of manned vessels too risky. During autonomous navigation ARCAB maintained a nearly constant speed of 1 m/s along survey lines, slowed during turns, and stopped at the end of its mission.
ARCAB performed well in both test deployments in Aarhus, Denmark in May 2018 and in southwest Greenland in January 2019. ARCAB demonstrated its stability and wave-piercing capabilities during the January 2019 deployment in Greenland. ARCAB navigated multiple autonomous routes within the specied boat speed and waypoint radius parameters. Overall, ARCAB satised all the design requirements.
ARCAB has been submitted to HardwareX and more expeditions are in the pipeline for 2019, enabling ARCAB to gather even more data, which can help the community understand climate change related problems.
Wow, I never had an autonomeous robot perform so well and problem free in Greenland. We had a saying, everything goes wrong in Greenland, but you broke that saying.
But that is not the only thing you broke, as I told you, this project broke the grading scale in my course. I simply could not give you a high enough grade, fitting to the quality and quantity of the work you did.
In some ways, this is one of the most difficult projects I have had students submit to NI Student Design Contest with autonomous navigation, obstacle avoidance (both small and large), drift correction, connection loss error handling and the list goes on.
I hope you receive the recognition, that I firmly believe you deserve.