DJI AV Rumors: It Is Already An Autonomous Vehicle Company

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

---

Reuters recently reported that DJI is building a team to work on autonomous vehicles, citing job postings and anonymous sources within the company.

DJI declined to comment, saying it had no new announcements to make, but its jobs page shows several automotive positions. The company is hiring people to work on automotive electronics, automotive software, and autonomous driving. Given the automotive nature of these positions, it’s hard to argue that they’d be for anything but ground-based autonomous vehicles.

Reuters’ sources said the company is building an automotive team and plans to sell driver assist technologies. While whole systems would be sold, it also plans to sell components, like lidar sensors.

Details are thin, but looking at what DJI is already doing in the space can shed a lot of light on the chances of DJI actually making this work. Truth be told, it is already an autonomous vehicle company, so it’s not a stretch at all.

What My Mavic 2 Can Do

To see what I’m talking about, all we need to do is look at its current products. I’ll use my Mavic 2 Pro as an example. It’s a drone with a very decent camera for the price point, but it also has 8 other cameras. It has two on the back, front, and bottom, and one on each side. It also has bright LED lights and an infrared sensor on the bottom, and an infrared sensor on the top, pointing up. Together, these cameras and sensors give my drone a decent obstacle avoidance system.

I’m careful to not test the limits, but it has refused to go closer to a tree on one flight and it always self-lands when I bring it in at the end of a flight. It also has a display overlay on the controller that tells you how far away any obstacles are, so it’s definitely “aware” of its surroundings.

It also has other intelligent systems. It uses its GPS sensor, magnetometer (compass), and connection to the internet (via my phone, attached to the controller) to always show me my position on a map, or relative to my position as pilot. I can also pre-plan flights on my phone or computer, giving it a series of waypoints (latitude/longitude plus altitude above ground level) and instructions of what to do at each point from takeoff to landing.

According to FAA rules, I’m still responsible for what it does and it must stay within line of sight, but the drone is capable of doing a lot more than pilots are legally allowed to do with it. The reason we can’t send it out for solo flights and/or miles away at this point is safety, especially when it comes to the possibility of collisions with other aircraft.

To help address safety issues, the drone is also programmed with “no fly zones.” Airspace near airports, military bases, high-level officials like the President of the US, and even wildfires are listed in the drone’s map as off limits or restricted by height. In some cases, you can send DJI a request to unlock an area if you can show you got authorization to fly there.

In the event I lose controller signal, the drone will come home. Older models would fly almost straight home if signal was lost, regardless of what might be in the way. The Mavic 2 drones, assuming everything works right, are “obstacle aware,” and will go over or around an obstacle on the way back home to avoid hurting themselves, people, or property.

Putting All This Together

I know the Mavic’s capabilities might not seem like much compared to driving, but that’s only if you don’t consider the complexities involved in operating small autonomous aircraft. Instead of largely operating in two dimensions, a drone must operate in three (back/forth, left/right, up/down). Trees, buildings, power lines, construction cranes, birds, helicopters, and airplanes all share that space with it. There are no “rules of the road” to count on with some of the other moving obstacles, especially those birds, who don’t know what government even is and may actually be out to get you.

The legal landscape can be complicated, too. Don’t believe me? Just look at the chart above showing what the rules are in the Dallas/Ft. Worth area. To get my license, I had to study what all this means, but there is a key on the edge of the chart or on another page of the test book to help you figure it all out. There are also decent apps that show a simplified version with only data pertinent to drone operations, but you still need to know the basic rules, and even then there are sometimes head-scratchers.

There aren’t any lane lines, curbs, signs, or traffic lights in the sky. There are rules, but they aren’t always obvious. Edge cases are everywhere.

Weather is also another important consideration. A 25 MPH wind won’t blow a Tesla around much, but it could make you lose control of a smaller DJI drone, and the weather can vary depending on altitude. Just 100-400 feet in the sky, things can be a lot nastier than they are on the ground. A drone pilot needs to keep track of all this (apps can help here, too), but the drone itself needs to be able to use GPS to keep the drone in one spot or going on a straight line despite the breezes and sudden gusts trying to move it around.

The bottom line is that DJI’s drones are already capable of doing some autonomous operations in areas more complicated than the roads, and sometimes far more complicated. It’s not a major stretch to say that DJI is already pretty good at dealing with complicated and dangerous conditions with its current autonomous vehicle efforts. Adding ground-based systems to its portfolio isn’t that big of a leap.

It also presents good opportunities for the company. Autonomous driving is going to be an area with a lot of growth in the coming years, especially in the Chinese market. Beyond that, the company already has a global sales organization and a global supply chain.

For all of these reasons, I wouldn’t laugh at DJI’s efforts. It is already into autonomous vehicles, so it’s geared up to become a decent competitor in the space.