Year submitted: 2019
University: Aarhus University, School of Engineering
List of Team Members:
Jonathan Pasma, 2019
Mathias Edslev Jacobsen, 2019
Main contact Email address:
Title: ISeDD – Iceberg Sediment Drilling Drone
Description: ISeDD is a research device which has the objective of collecting data from icebergs in the fjords of Greenland. ISeDD 1.0 is primarily an ice drilling drone, which means that it is capable of extracting an ice sample from an iceberg. Furthermore, ISeDD also has the ability to deliver a GPS beacon onto the iceberg, which enables researchers to track the icebergs that the samples come from. This data is needed for determining where, how much and what kind of sediment the icebergs releases into the fjords, which all have a large impact on the local environments in the fjords and the northern part of the Atlantic Ocean.
List of the used products for building the ISeDD 1.0 prototype.
NI LabVIEW 2018
NI MultiSim 14.1
NI LabVIEW 2018 Real-Time Module
NI LabVIEW 2018 myRIO Toolkit
DJI Assistant 2
- DJI Matrice 600 PRO
- DJI controller with CrystalSky monitor and Channel expansion kit
- DJI Zenmuse x3
- Self-developed ice cup drill with brushless motor and gearing
- Self-developed GPS beacon release
- Self-developed linear actuated landing gear with stepper motors and leadscrews, as well as floats for emergency water landings
This project was built around a problem given by Daniel Frazier Carlson (Dan Carlson) from the Institute of Bioscience, Arctic Research Centre (ARC) at Aarhus University (AU). Dan Carlson is an oceanographer and specializes in ice/ocean interactions in Greenland.
The knowledge of understanding how icebergs melt and how much freshwater they release into the sea is of great importance to the climate. The studies include melt speed, drift speed, the amount of freshwater produced and what materials/sediment the melt contains, as all of these factors will help researchers understand how much icebergs influence the climate. A reduced amount of icebergs will lower the concentration of sediment and is expected to play a huge role in the biology in the water of the fjords. Especially phytoplankton which lives off the minerals in the sediment can be influenced, which will result in large changes to CO2 absorption as well as decreased supply of food for other water species. The ice-sheet/glacier ice on Greenland is affected by the climate changes, and therefore it is more important than ever to analyse the effects from the icebergs.
These studies are made difficult though because of the nature of icebergs; the arctic environment, hard to reach places and dangerous conditions when working on or close to the icebergs. It is not suited for a person to collect data from icebergs for several reasons. The fact that the icebergs are often inaccessible will limit the number of icebergs that can be tested. Furthermore, icebergs have a high tendency to roll over as they melt, which can endanger the person working the drill, who risks falling into the ice-cold water. The wave from the roll can also cause the boat to flip over, endangering the whole crew and the equipment. In other words, it is hard to collect the important data needed for the studies.
The data can be used to understand the ice and water conditions in the fjords, but also in advanced models of the sea, since a lot of the icebergs continue out of the fjords and into the Atlantic Ocean to melt. Therefore, the melt plays a significant role in the amount of freshwater and the temperature of the Atlantic Ocean, which has a large impact on the global climate conditions.
As a helicopter is too expensive to use, Dan Carlson made a project suggestion of a drone collecting the data and proposed the two most relevant data sets that is needed from icebergs, which are described underneath.
As an iceberg melts it leaves sediment in its wake. By researching what types of particles, the melt contains and by which icebergs it comes from, it might be possible to find a correlation between iceberg drift and the local biology as well as the environment.
The sediment on an iceberg is made up of sand, pebbles, rocks and micro particles and it might even contain trace amounts of iron, from the bedrock underneath the glaciers. If the sediment of a melting iceberg can be analysed and scientists are able find the amount, the composition and the size distribution of the sediment, it will be possible to understand how the sediment dissipates in the water, and what materials are led into the fjords. The target was to retrieve ice sample of at least 10 cm in depth and 500-1000g in weight to analyse.
Other than contributing to the environment in the fjords, the sediment also changes the reflection of sunlight on the iceberg surface, also known as “Albedo”. This has never been accounted for in models of icebergs, which is another important feature that the drone needs to have.
Another important information to know about the icebergs is how they travel once they are released from the glaciers into the fjords. The sediments from the iceberg is transported via the melt into the water. Where this sediment ends up, can be analysed by tracking the icebergs position with a GPS beacon. Through GPS data analysis it is possible to the determine how the majority of the icebergs move, and how the distribution of sediment is expected to be.
After spending almost half a year on design, construction and testing, the solution is our working prototype: ISeDD 1.0. The device is an ice sampling drone built around a DJI Matrice 600 Pro frame, which has been modified to carry a developed ice cup drill, a GPS beacon release and redesigned landing gear for operating the drills movement up and down. The drone has an electronics compartment fitted underneath the main frame which carries the NI myRIO, the main circuit board, the stepper drivers for the landing gear, the ESC for the brushless drill motor and the drill motor and gearing itself.
To start the design process, first a complete CAD assembly of the Matrice 600 Pro was constructed, so the parts could be designed and fitted properly. The final CAD assembly of ISeDD 1.0 can be seen underneath, with both the drill and GPS beacon fitted. As an added safety feature, a Roll alarm LED has been added in the field of view of the drone’s camera. This alarms the operator if the iceberg is about to roll over, as it measures the change in angle compared to gravity, with zero being the angle just after landing.
The electronics compartment is placed underneath the main frame and uses the original mounting points for DJI’s own products (the 4 black rods with triangular mounts on top). The compartment is built on a CNC mounting plate with ribs for added stiffness. The myRIO is placed in a custom 3D printed mount and held in place with a Velcro strap. This Velcro strap also secures the separate battery for the drill which is not shown in the picture. The main circuit board is placed at the other side of the plate, with the drill motor and ESC in the middle. The DJI Zenmuse X3 camera is mounted under the triangular plate, with a cover for the wires. At the other end, the mounting plate is ready for a small GoPro sized camera to be mounted. This camera enables the drone to measure surface reflection (Albedo) with a light calibrated camera. The drill is mounted underneath the mounting plate in a CNC milled bearing house and held in place with a connector to the motor axle. Next to the drill, the GPS beacon release is mounted holding the GPS beacon in a string.
Main circuit board:
In order to have an organized electronics compartment and have easy access to plug in the steppers, end stops, fans, GPS beacon release, the Roll alarm LED and the in- and outputs from the drone and myRIO a main circuit board has been made. The circuit board acts like a motherboard for the drone and controls all the functions of the drone. The power is supplied from the drone’s 18V output and converted down to 12V. The board receives 5 inputs from the drone’s flight controller, which is then send to the myRIO and processed in the VI’s. After processing the myRIO sends out the orders and enables the requested function.
The construction of the motherboard was made easy by making the entire circuit in NI MultiSim 14.1, in order to secure that everything was connected properly, see the final schematic underneath.
The tasks of delivering a GPS beacon onto the iceberg is solved by using a linear solenoid, with a push spring. This means that the axle from the solenoid can be used as a hook, which is always closed. The GPS beacon is transported in a 2m rope and released safely onto the iceberg by powering on the solenoid which releases the rope. In the picture below, the solenoid mechanism can be seen with a red rope for the GPS. The dropping operation can be monitored with the onboard camera.
Roll alarm LED:
The Roll alarm LED is placed in the field of view of the DJI Zenmuse X3 camera, and shines light into the camera to alert the operator. The alarm is needed as the melting icebergs can roll over several times a day. The function utilizes the accelerometer on the myRIO, and triggers if the angle is above a certain threshold.
During the development a lot of time was spend on designing the ice core drill, as ice is an unpredictable material. The solution ended up with a modular drill, with an aluminium center core, an aluminium drill top, a 3D printed PLA sleeve for transporting material and 3D printed PLA drill head. The 3D print made it possible to do rapid prototyping in the design phase and was thoroughly tested for durability with excellent results.
The drill was fitted with a center drill for added stability during start-up, which was a significant improvement. Furthermore, the drill is fitted with 3 hardened stainless steel teeth, with different cutting surfaces for reducing the size of the ice shavings when drilling, see picture below. The center drill and cutting teeth were PVD coated for added abrasion resistance.
During tests it was clear that the ice core had a high tendency to break off when the mount from the center drill hit the ice, which became the concept for snapping off the ice samples. The drill then relies on friction to pick up the sample after it has broken off.
ISeDD main parts list:
Underneath follows a list of the different parts on ISeDD 1.0.
As flying a drone autonomously is difficult to get approved legally, it was decided during the project that the pilot should be able to fly and control the drilling process manually using only the controller. Therefore, a channel expansion kit was fitted on the DJI controller. This enabled the feature of accessing the PWM outputs on DJI flight controller and make the controller the user interface, which can be seen in the picture below.
The software for the drone’s functions is made up by a system of parallel while loops with the main loop being a state-machine. The purpose of the state-machine is to control the movement of the landing gear stepper motors. The state-machine receives 2 inputs from the expansion kit interface, as well as a grounded signal which is toggled when the drone is landed (“Grounded”). Once the drone is “grounded” it is possible to activate the step drilling, which is the state shown in the picture underneath. The step drilling cycles between 14s downwards movement and 0,8s upwards. This process has been tested thoroughly and the upwards movement was added to secure reliable drilling. It was found out that the ice building up on the cutting teeth can cause stress on the drill which can make it stop, therefore the upwards movement was needed. The state-machine also checks if the end stops located on the landing gear are triggered. Besides the state-machine the other while loops are used to control the other aspects of the drone such as, starting the drill, dropping the GPS beacon and checking if the iceberg is tilting. All functions are made with low level components in order to secure a fast execution.
Field test Greenland:
ISeDD 1.0 has been tested in Greenland from the 5.-11. of January 2019, with a successful result. The mission was carried out from a small modified boat, where the drone could be operated from a small 1,6x1,6m plywood platform for take-off and landing, see picture below.
The drilling capabilities was tested on easily accessible sea-ice and showed that the developed drill and landing gear, was very good at drilling in Greenlandic ice. This was very encouraging before going out in the fjords and drilling in icebergs.
The drone was tested on 2 icebergs with a successful extraction from the iceberg shown underneath (25m long and 12m high).
The extracted ice sample can be seen in the photo below. The drill is easily dismounted from the drone and together with a small extraction tool, the ice sample can be removed. The sample was put in a sealed bag directly after being extracted from the drill, in order not to contaminate it. The sample is currently stored in Nuuk, Greenland and will be analysed in the spring of 2019.
The benefits of using LabVIEW and NI tools:
The brain of the prototype is a NI myRIO which is a powerful tool in combination with LabVIEW. As the drone’s flight controller was fitted with the expansion kit, it was possible to receive the user input directly on the myRIO, and at the same time having it execute the right commands. The programming language in LabVIEW is very intuitive and it is easy to build large programs with many in- and outputs, which was a necessity for this project. Furthermore, the capability of easily making parallel loops which can be processed on the Dual core myRIO, speeds up the whole program. As the NI product line consists of many devices that work well together, and the fact that NI products and LabVIEW is part of the Mechanical Engineering courses at Aarhus University, it was an obvious and good choice for this project.
Level of completion (beta, alpha, or fully functional):
The ISeDD drone is a fully functional prototype which was tested in Nuuk, Greenland in January 2019, as mentioned earlier. The drone succeeded in collecting ice from an iceberg, which was a huge success on its first field test. The GPS beacon released has only been tested in Denmark, as there were no suitable icebergs to track at wintertime in Greenland. This means that the drone is fully functional, and capable of fulfilling its mission.
ISeDD 1.0 is in general a success as it was able to deliver an ice sample in Greenland, which is “proof of concept”. The drone can be used to collect important data for iceberg research but can also be used in other areas. One of these areas would be thin sea-ice as no human can walk onto the thin ice and collect data, as there is a danger of the ice breaking. The thin sea-ice holds a lot of information about mercury and boron concentration which is led into the atmosphere and is another use-case for the drone. Lastly the drone can also be used on glaciers with hard to reach places. These things both applies to research in Greenland but also other arctic areas such as Norway, Iceland or even Antarctica, which means that the data from the drone can have a huge impact for many scientist’s research worldwide.
The development of ISeDD continues at Aarhus School of Engineering and is currently in the process of getting its drill capabilities upgraded, with a semi-autonomous drilling feature and better ability to collect the sample. This means the project will be improved and tested further in the future, in order to have the best possible device for collecting ice samples, for such an important scientific research area.
As of February 2019, Dan Carlson is in the process of publishing an article in “Frontiers” magazine about the drone and its impact, which will be submitted in the spring of 2019.
Furthermore, the development process and the field test to Greenland was accompanied by a documentary film crew for the Danish National Television, who are interested in portraying the project as part of a documentary. The documentary is set to be released in 2020 and called “Into the ice”.
Time to build:
The project was carried out in the fall of 2018, from mid-August to December, with fieldtrip in January 2019. The project consisted of roughly 2 months of designing and 3 months of testing and constructing. It was made as a Bachelor Thesis, with a time estimate of roughly 500+ hours per group member.
Staring in a poster:
ISeDD 1.0 features in a poster for Aarhus University School of Engineering in the spring of 2019 (in Danish: “Do as 4000 others: Study Engineering in Aarhus”).