The Skydio R1 is not a perfect drone. It produces 4K material, but the image quality is paler compared to DJI drones. It's relatively large, it does not work, and at $ 2,500 it's way too expensive for … basically anyone. But today, the small Californian startup adds an update to R1's flagship autonomous flying capabilities that makes this price tag a little less painful.
Especially the first major software update from Skydio offers the opportunity to follow vehicles. The company has been training the neural networks running on the drone's embedded computer to detect everything from cars to 4
In my view, this is exactly the kind of thing GoPro should focus on from the start with karma, although admittedly it's just one of several reasons why karma failed.
Think back to one of the first teaser videos from GoPro Karma. A skier rushes to a jump and holds a camera in selfie position all the way. When he beats his lip, he lets go of the camera, and she glides gently into the sky, offering a magnified view of him and his friends as they descend the mountain.
The execution of this moment is more complicated than in the display. GoPro only added a "Follow Mode" in September 2017, just four months before the company completely removed its drone business. That is, if the skier let go of the drone, someone else held the controller in hand, ready to carefully fly the karma away from the humans below. This person also had to be hidden somewhere (or maybe even removed digitally from the shot).
Anyway, this ad made it look like Karma users effortlessly blow up the drone could throw them on their adventures. Other teaser videos have also taken up this idea, but with less dexterity. GoPro played out the idea of a drone that can be "thrown in and kicked off" that has survived since the advent of the market. (Do you remember Lily?) But only with Skydio did a company really deliver.
It did not take more than a year for GoPro to add a feature similar to what Skydio now offers, Karma's "Follow Me" mode was a shadow of the feature when we thought about it. GoPro did not do any computer vision on Karma to run this mode, like Skydios or DJIs drones. Instead, GoPro simply had the drone pointed in the direction of the GPS signal coming from the controller.
In addition, karma did not have any of the obstacle avoidance features that combine hand and hand with autonomous flight. Drones can not fly themselves without knowing where they are and what is around them, and Karma has suffered this inherent problem. It had no cameras other than the GoPro that you attached to the gimbal. Meanwhile, the R1 has 13, plus motion and depth sensors and an Nvidia chip, which handles the autonomous processing.
In this light, it's no wonder that autonomy is the thing that sets the Skydio drone apart. It is explicitly designed to be a drone that can follow you anywhere and navigate between obstacles or around. Not even DJIs drones are so autonomous. Many of them (apart from the 499 Spark) provide object tracking and can follow people or things in open spaces. But while DJIs drones are able to avoid hitting a tree or building while following you, they just do not offer the same promise of full autonomy as Skydio with the R1.
I'm not saying that GoPro should have followed exactly what Skydio did with the R1. There are reasons to take a different path. First, deliberate build-up of autonomy would probably not have sustained karma any longer, because for all the reckless actions they normally capture, GoPros are often used with care by professionals and semi-professionals alike. This means that manual control must also be in front and in the middle. But the idea of a drone that can follow you when you're tearing down a mountain or circumnavigating a circuit – with no one flying away – is just as relevant to the GoPro brand as the raging Red Bull.
There's less adrenaline as well causing money problem – specifically how much GoPro was made (or not made) on every karma sold. The cost to make karma, as we know it, combined with DJI's dominance of the market, were key factors in the decision to discontinue the drone, said GoPro CEO Nick Woodman The Verge earlier this year ,
"We looked at how much we spent on our drone program, in proportion to the number of units we sold, and above all the profit we made for the whole program," he said in January the CES. "It became clear that the drone category will continue to be a low-margin category, with incredibly strong competition."
The US commercial drone market is still growing, according to the latest data from the NPD Group, but most of the growth is coming currently from drones that cost between $ 500 and $ 999. If this trend continues, GoPro, which has opted for something more premium and technologically advanced like the R1, would not have changed much, unless the company was really ready to take a dip on the product.
In addition, GoPro's business depends entirely on volume. In its most recent earnings report, the company noted that it was the top-selling camera in the entire digital imaging category for the [17April1969] in North America . In the first quarter of 2018, it was 86 percent of the market for action cameras.
Considering how expensive the R1 is, it is believed that even GoPro could have been difficult, with all its manufacturing efficiencies, to do something similar for the "at or below $ 1,000" price tag that is obviously required to make a significant profit.
Regardless, the gap between what the R1 and what Karma can do when it comes to throwing and droning drones is wider than any gap you could dare to make with a dirt bike. And it will continue to grow as Skydio delivers more updates like today. The startup may never come so far that its drone is as "cheap" as Karma, or match the image quality of GoPro or DJI. It's only a matter of time before DJI connects Skydio with his own completely autonomous flying, and that's what the startup has to expect. But even if it ultimately fails, the Skydio R1 remains a glimpse of what the karma should have been.