Mavic Air; See and Avoid

Assignment 7.4
For the past 10 years, DJI has been leading the market in consumer drone sales (Glasser, 2017). Their technology is evolving with every product release and their newest model, the Mavic Air, proves to be one of their smartest drones yet. Many new DJI owners quickly find out how difficult it is to fly a drone, and accidentally crash into an object. DJI has begun integrating sense and avoid technology into their aircraft to help keep many of these crashes from happening.  
The Mavic Air boasts 7 cameras and 2 infrared sensors positioned around the aircraft to give it a better situational awareness of its surroundings. The primary camera captures 12mp stills and 4k resolution video, which is the primary reason somebody would purchase the Mavic Air, however, this high-resolution camera is also tied into the aircraft's navigational system (Horaczek, 2018). The other  6 cameras are positioned in pairs around the aircraft with some separation between them to provide the necessary parallax angle to view things in 3 dimensions (Horaczek, 2018).
All of the data absorbed by these sensors needs to be processed to be turned into usable information for the aircraft. Additionally, this needs to happen fast enough for the aircraft to respond in time to avoid a crash. “In addition to the hardware upgrades, DJI has also improved its algorithms for crunching all that data it’s sucking in. For instance, instead of stopping when the drone sees an object in its path, it will now examine the 3D map of the terrain it created and look for a clear path around the obstacle” (Horaczek, 2018). DJI calls this their Advanced Pilot Assistance System (APAS). “Users can keep flying forwards and backwards, and APAS will plan an appropriate flight path to bypass obstacles automatically” (DJI, 2018). Having the ability to create a 3d map in real time flight is critical in avoiding objects around it. This same method of developing a 3d map for seeing and avoiding is used in self-driving cars. It is important to consider how the available power for flight and endurance is impacted by powering the see and avoid systems. As additional power is required for sensors, the endurance of the aircraft will be decreased.
The integration of additional sensors and advanced algorithms have made the Mavic Air very capable of seeing and avoiding most objects that would be in its flight path. The system's shortcomings are evident when the aircraft is moving from side to side, as there are no sensors pointing to the left or right. Future aircraft could benefit from the existing sensor layout of the Mavic Air, with the addition of side looking cameras. This addition would likely require a more powerful processing unit thank currently exists on the Mavic Air.
                The Mavic Air is an excellent example of a consumer-grade unmanned system with advanced see and avoid capabilities. As this technology continues to evolve, unmanned systems will become safer for the operators, and their surroundings.

References
DJI. (2018). Mavic Air: Adventure Unfolds. Retrieved from https://www.dji.com/mavic-air
Horaczek, S. (2018, January 23). DJI's Mavic Air Drone uses more than a dozen sensors to keep it from crashing. Retrieved from https://www.popsci.com/dji-mavic-air-sensors-object-avoidance

Glaser, A. (2017, April 14). DJI is running away with the drone market. Retrieved from https://www.recode.net/2017/4/14/14690576/drone-market-share-growth-charts-dji-forecast

Comments

Post a Comment

Popular posts from this blog

Unmanned Aerial Systems: DARPA, CODE and LAWS

Image
Unmanned Systems are nothing new in the Air Force. Remotely pilot’s vehicles have been at the forefront of the battlefield for decades now. But these systems all rely heavily on functioning communication links, and multiple crew members and support analysts. Without these key elements today’s drones are hardly capable of accomplishing any mission at all on their own. Operation in denied or contested airspace would be especially difficult even with the application of low observable technology.  The Defense Advanced Research Projects Agency (DARPA) has been working to add more autonomy and give existing and future systems the capability to operate in these denied and contested environments. CODE is DARPA’s answer to this problem. “DARPA’s Collaborative Operations in Denied Environment (CODE) program aims to overcome these limitations with new algorithms and software for existing unmanned aircraft that would extend mission capabilities and improve U.S. forces’ ability to conduct oper
Read more

OpenROV Data Display

Image
Open ROV VR Cockpit             In 201X the OpenROV company released their idea on Kickstarter for a DIY underwater drone to explore the ocean.  In an Interview with the company’s co-founder David Lang he stated “five years ago the team just wanted to explore underwater caves in search of gold, but in the time since then, they’ve shipped 1,600 underwater robot kits to 50 different countries, prototyped and experimented with numerous new designs, and built a community of similarly minded underwater enthusiasts online” (Love, 2015). Their campaign was a huge success and the evolution of the design began to grow in sophistication both in terms of performance, and available data. What makes the platform special is the fact that it is open source, which means its software and development is wholly supported by its community of users. One recent addition to the Open ROV platform is the OpenROV VR Cockpit which displays valuable data to the operator directly to a virtual display on the u
Read more

3.5 Research Assignment: UAS Sensor Placement

Image
Research Assignment: UAS Sensor Placement Preface, This topic is of particular interest to me as I own both of the UAS being discussed in this assignment/blog posting. The DJI Mavic Air, and the Armattan Rooster FPV racing/freestyle Quadcopter. First Person View Drone Racings is gaining rapid popularity thanks to the Drone Racing League being featured on ESPN, local racing events held by Multi GP, and freestyle flying as seen on the popular youtube channel Rotor Riot. The high speed quadcopters are built to be fast and durable typically touting carbon fiber frames, high thrust motors, and advanced flight controllers. The Armattan Rooster is a Racing/Freestyle frame that can be customized by the builder or purchased as a complete drone. My personal Rooster was built using a CLracingF4s Flight controller, and a 4 in 1 Spedix 30 am ESC, 4x TMotor F40 Pro v2s, a TBS unify pro video transmitter, and a TBS crossfire long range low latency receiver. The onboard FPV camera is a Runcam
Read more