Sense and Avoid Sensor Selection

Hi Everyone!

For this week's research assignment, I chose the DJI Phantom 4 to discuss its sensor suite for collision avoidance technology.

I chose this small UAV because it is considered the leader in commercial UAVs for aerial photography and video.  It's flagship model, the Phantom 4, was recently released for the mass market in Spring 2016 and it contains small improvements on speed and battery life over its predecessor but primarily boasts its novel sense and avoid system which separates it from the rest of its rival small UAVs.

DJI Phantom 4

The DJI Phantom 4 costs approximately $1399 USD.  Its take-off weight is 1380g, travels a max speed of 20 m/s, operates at a maximum ceiling of 6,000 metres and can last for up to 28 minutes of flight.  The Phantom 4 is powered on a single LiPo4S battery with 5350 mAh capacity and takes approximately 1 hour for a full charge at maximum power of 100W.

Dual Under-Carriage Sensors

The sense and avoid system have a range of up to 50 metres and has 60 degrees of viewing angle. It has two forward facing optical cameras and two sonar and optical cameras underneath for ground detection.  The minimum separation distance between obstacles is programmed at around 2 metres, so any intentional or accidental collision courses would be prevented.  The main limitation of this small UAV sensor system is the lack of overhead sensors and rear facing sensors.

Dual Forward Facing Sensors

The technical process that governs the entire sense and avoid system is focused around its flight controller which acts as the brain to compute and analyze all of the inputs gathered from its front and under-carriage sensors.  The computer algorithm calculates incorrect data and invalidates it to determine obstacles, then cross-references it with its positional data gathered by its proprioceptive sensors such as its dual INUs for inertial status and its barometer for altitude reading.  The GPS also acts to determine its position relative to its environmental settings and the flight controller outputs the requisite manipulations to coordinate adjustments to its 4 motors, resulting in a mid-air side-stepping of any potential obstacles.

Phantom 4 Sense and Avoid Path

Although expensive, this compact sUAS is the leader in aerial photography drones currently available in the market.  It is a leader in sense and avoid technology and boasts the follow-me function without the operator having to carry and transponder device.  It is a great asset whose collision avoidance technology could be leveraged for other unmanned systems.

6.5 - Control Station Analysis

Hi Everyone,

For this week's assignment, I chose the unmanned underwater vehicle called the Crabster CR200.  It is a remotely operated vessel that was inspired by the the natural design of lobsters and crabs with a primary objective of being able to steady itself for deep seafloor explorations.  Past vessels that wanted to explore the ocean floor were ill-equipped to brace and counter the harsh undercurrents and turbulent deep-water tides, as such, out of this necessity, Crabster CR200 was born.

Crabster CR200

The remote control unit that operates the Crabster is housed at the surface level within a 20 ft sea container.  Within the sea container, it contains a lot of equipment that is meant for a team of four operators to use.  You have the pilot, co-pilot, navigator and also sonar/sensor operator - it almost looks like an aircraft cockpit!

Multi-Program Driven

The sea container consists of seven computers, nine LCD monitors, four joysticks, and power supply.  The hardware is powered on a Intel Quad Core i7 CPU, 8 GB main memory, and Gigabit Ethernet interface.  The agent program is based on Linux Software but there are also many commercial off the shelf (COTS) software that is used such as the Navigation Program and Video Program which are all integrated via the Agent Program.

The data presentation of the Crabster is mostly traditional 2D views and live-video views from the pilot view.  However, the pilot video does have an interface similar to a heads-up display found (shows tide/current speed, pitch/roll/heading and speed) on most modern aircrafts which helps, from a human factors perspective, assimilate information much more easily.  There is also a 3D-view of the Crabster itself that is made possible by the on-board sensors relaying positional and other environmental state information back to the remote control station which then is simulate through the 3-D viewer program to create a dashboard view of all critical information of the vessel itself.

3D Dashboard View

Pilot HUD view

For future improvements, it must be stated that the portability of the control station needs to be investigated to ensure its footprint can be made more concise.  It is currently cumbersome to transport and requires lifts and transportation capable to move a sea-container.  That said, I do believe the 4 personnel team required to operate the Crabster CR200 is a good design choice as it allows each specialist to focus on their task at hand to optimize results and share the overall workload of such a complex unmanned underwater vehicle.

Remote Control Station

Hope you enjoyed my research for this week - cheers!

References

IEEE Spectrum. (2014, July 30). IEEE Spectrum. Retrieved from Huge Six-Legged Robot Crabster Goes Swimming: http://spectrum.ieee.org/automaton/robotics/industrial-robots/six-legged-underwater-robot-crabster

Kim, B., Shim, H., Yoo, S.-Y., Jun, B.-H., Park, S.-W., & Lee, P.-M. (2013). Operating Software for a Multi-legged Subsea Robot CR200. Retrieved from IEEE Xplore: http://ieeexplore.ieee.org.ezproxy.libproxy.db.erau.edu/stamp/stamp.jsp?tp=&arnumber=6608151

Sea Technology Magazine. (2013, October 13). Hexapod Robot Crabster CR200 for High Tide, Turbid Water Exploration. Retrieved from Sea Technology Magazine: http://www.sea-technology.com/features/2013/1013/6.php