Air 3S: Advanced RTH in Complex Environments with LiDAR

[msinger]

The DJI Air 3S enhances the safety of its Return to Home (RTH) feature with forward-facing LiDAR, allowing it to detect obstacles even in low-light environments. This omnidirectional obstacle sensing helps the drone avoid obstacles like buildings during its flight and return paths, improving safety during night flights in urban areas. Additionally, it can memorize flight paths in well-lit conditions, allowing for quick and safe returns even when satellite signals are unavailable.

In this video, Air Photography demonstrates the impressive capabilities of the Air 3S as it navigates back to its home point in a tree-covered environment.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[wco81]

Saw a clip where the aircraft at night or in twilight is flying about 20-30 feet high towards a lit apartment balcony, goes right through the glass doors and lands in a table inside.

So you don't need to worry about setting a safe RTH height?

Or was it flying at 50 feet or higher and then for the last 100 feet or so it dropped in altitude so that it was level with the balcony doors, ending up like 3-4 foot above the table inside before finally landing on it?

Is the lidar definitely going to detect power lines or other thin profile obstacles which the drone might encounter in its RTH path?

 

[msinger]

Is the lidar definitely going to detect power lines or other thin profile obstacles which the drone might encounter in its RTH path?

Like any automated features in DJI drones, you should use them with care like any tool in your toolbelt -- rather than a safety net that is always there to catch you.

Here's some guidance from the Air 3S manual:

 

[wco81]

Yeah that video was mocked up for marketing.

And the manual has CYA verbiage, though basically saying that the kind of operation depicted is not supported or even recommended.

 

[msinger]

that video was mocked up for marketing

I think it's a great example of what the Air 3S is capable of when flying in good conditions.

And for DJI's recommendations, they are always going to tell people to take the safest route. Because that's the most successful way to fly safely.

 

[Hill Peaceman]

The DJI Air 3S enhances the safety of its Return to Home (RTH) feature with forward-facing LiDAR, allowing it to detect obstacles even in low-light environments. This omnidirectional obstacle sensing helps the drone avoid obstacles like buildings during its flight and return paths, improving safety during night flights in urban areas. Additionally, it can memorize flight paths in well-lit conditions, allowing for quick and safe returns even when satellite signals are unavailable.

In this video, Air Photography demonstrates the impressive capabilities of the Air 3S as it navigates back to its home point in a tree-covered environment.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Unfortunately no support of SDKs

 

[msinger]

Unfortunately no support of SDKs

SDKs for what?

 

[GFields]

Awesome, thanks for sharing.

 

[Hill Peaceman]

SDKs for what?

For planning your flights and use photogrammetric software

 

[msinger]

For planning your flights and use photogrammetric software

Ah, right. That's an entirely different topic though.

 

[Droning on and on...]

Is the lidar definitely going to detect power lines or other thin profile obstacles which the drone might encounter in its RTH path?

All sensing systems have limitations, including your wet, squishy spherical pair

We may have crossed the point now where, with Lidar, the Air 3S can see and discriminate thin obstacles better than humans under certain conditions.

Regardless, there will be situations where the new sensing system will fail. Push it to limits or beyond, and like car brakes, it may not save you.

 

[Droning on and on...]

The DJI Air 3S enhances the safety of its Return to Home (RTH) feature with forward-facing LiDAR, allowing it to detect obstacles even in low-light environments. This omnidirectional obstacle sensing helps the drone avoid obstacles like buildings during its flight and return paths, improving safety during night flights in urban areas. Additionally, it can memorize flight paths in well-lit conditions, allowing for quick and safe returns even when satellite signals are unavailable.

In this video, Air Photography demonstrates the impressive capabilities of the Air 3S as it navigates back to its home point in a tree-covered environment.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

For those of us that have been flying DJI drones for many model generations, this was fall-on-the-floor, catch-your-breath amazing.

Now I gotta have one for no other reason. Going to order right now.

 

[Droning on and on...]

Yeah that video was mocked up for marketing.

Evidence?

Looked uncut during the most important part.

 

[Droning on and on...]

For planning your flights and use photogrammetric software

While a bit cumbersome, PG flights can be planned with other, non-DJI tools, then converted to DJI waypoint format, loaded into Fly, then flown with the native DJI waypoint feature.

An SDK would make all this easier for the end-user, but isn't necessary to accomplish things like grid flights, if this drone has unique features that would be helpful.

For automated flights, the greatly improved OA is, IMO, a huge benefit.

 

[wco81]

Evidence?

Looked uncut during the most important part.

I'm saying it shows a scenario which most people wouldn't try and is likely to draw a lot of attention, flying a drone at night low through a neighborhood like that.

And if there's damage to the drone or property, DJI would probably say it's not a supported operation.

 

[Droning on and on...]

I'm saying it shows a scenario which most people wouldn't try and is likely to draw a lot of attention, flying a drone at night low through a neighborhood like that.

Okay, that's fair.

I have a very different definition of "mocked up"... It means "fake" to me.

 

[Hill Peaceman]

Ah, right. That's an entirely different topic though.

But Your demo is impressive and if this drone was supported I will surely buy one.

 

[msinger]

But Your demo is impressive

That demo was from Air Photography. He has a lot of informative videos on his YouTube channel.

 

[wco81]

As impressive as it was, it would be useful if it could RTH at 50 meters, flying above the buildings, and then dropping down to the balcony where the Home Point is, then fly in 5-10 feet and then land in the balcony, which may have a ceiling or overhang.

That would minimize attracting attention.

The video where it's flying between buildings and flying maybe 1 or 2 stories above street level, over roads, for a couple of hundred feet, would be more likely to draw negative feedback.

 

[davidarmenb]

As impressive as it was, it would be useful if it could RTH at 50 meters, flying above the buildings, and then dropping down to the balcony where the Home Point is, then fly in 5-10 feet and then land in the balcony, which may have a ceiling or overhang.

I believe the Air 3S already does do this, if you take off from that balcony and have that spot as your home point.