Hello Tello Pilot!
Join our DJI Tello community & remove this banner.
Sign up

Windows 10 UWP app development

Joined
Oct 27, 2018
Messages
8
Reaction score
1
Age
36
Location
Cúcuta, Colombia
I just got a Tello and been unable to use it daily, since I use a Windows Mobile device, I know many people tells me to switch to iOS or Android at least for now, that's not going to happen. So I bought the Tello to be able to develop a Windows Mobile app to use it with my device, but also though about developing this app as a UWP (Universal Windows Platform) app that can run not only in mobile devices, but also in tablets, PCs, hololens and even Xbox. I wonder if anyone has ever heard of anyone already doing this kind of development so I can either base upon it or collaborate with it, because I think DJI could even benefit of having a complete suite of apps, for different drones or a single one that suits all, that can work in fully fledged computers for a better flying experience, and in some cases offer more control features that and iOS or Android devices. This way hardcore pilots can use their laptops with far better antenas, better screen and even better controls available when flying their drones for profesional usage. So there it is folks hopefully someone can help me with some information about it, otherwise I'll have to develop this whole thing by myself and hope for DJI to be interested, since I've noticed there are none apps remaining in the Windows Store for the drones already flying around. Thanks
 

Rat_rulz

New member
Joined
Jan 1, 2019
Messages
2
Reaction score
0
Hi David, very interested in this comment! And wondering if you had any luck either finding a UWP to test out or have started the process to use with multiple devices for the large screen and fimiliiar controls for ease of use for everyone.

Look forward to your response
 
Joined
Oct 27, 2018
Messages
8
Reaction score
1
Age
36
Location
Cúcuta, Colombia
Hi, currently I'm on an alpha development state, so no beta versions for now. But this app I'm developing is aimed towards Windows 10 as a Store App, which can be updated with no problem. My main objective with the app if to offer safety for the drone and people around it by, first of all, letting the app do all the connectivity stuff and make everything needed to keep the device with the drone connected, ensuring the safety again of the drone and people preventing to loose connection at hopefully any given time. Instead of relying on the Android or iOS connectivity functionality. Second, I'm trying to find a way to let this app update the drone firmware, so you don't need either an Android or iOS device to do so. Third, since I'm using a Windows Mobile device, my aim is to keep using it for as long as I can, and since it can work with Continnum I'm developing the app as a UWP version that can work with the device in Continnum mode as well as any other PC. This opens a wide variety of options, I guess most people would fly it with laptops, bigger screens, therefore more real estate for additional instrumentation, maps, controls. Also I want to achieve the same as the TelloFPV app that can work with mostly any Bluetooth control, including the GameSir T1D controller, making it easy to fly either by Bluetooth or with an USB OTG cable. And finally to make the whole UDP package transference for efficient, since most people send bursts of commands to ensure they are received, but I've noticed they tend to warm up the drone. So I also want the app to be "intelligent" in a way it can send as much and the least packages needed every time to make sure the command is received and at the same time the drone can both keep as fresh it can and save as much battery in the process. I guess with that flight efficiency measures I can squeeze a minute or two of additional flight time. So I still have a long way to go before I can release a beta test request, until I make sure the app is safe enough to be used in open space where things can go wrong, but still looking forward to prevent them from happening. So keep tuned and hopefully in a couple of weeks I might be asking for people to add them as beta testers in the Store and give the app a chance.
 

volate!lo

Well-known member
Joined
Dec 13, 2018
Messages
1,725
Reaction score
1,634
Location
Germany
I guess with that flight efficiency measures I can squeeze a minute or two of additional flight time.
I would be _very_ surprised if you could squeeze out even 20 seconds extra flight time by improving data transfer efficiency (unless you found a way for hacking the firmware). Optimizing a few packages will reduce Tello's power requirements by a tiny tiny fraction.
So don't spend too much time on optimization early in the process.
 
Joined
Oct 27, 2018
Messages
8
Reaction score
1
Age
36
Location
Cúcuta, Colombia
I would be _very_ surprised if you could squeeze out even 20 seconds extra flight time by improving data transfer efficiency (unless you found a way for hacking the firmware). Optimizing a few packages will reduce Tello's power requirements by a tiny tiny fraction.
So don't spend too much time on optimization early in the process.
I've been able to fly an additional 27 secs compared with the official app which in Android is the most efficient yet, total time 12:39. Yours is not so behind with a difference of 33 secs and the other aTello and aTelloPilot apps have a 40+ difference. I'm not currently being so aggressive in the package burst saving so I think that by optimizing it I could get between 20-30 seconds. But so far I've been testing with the Xiaomi Mi2 repeater, I believe it currently is the most popular one among the pilots. I need to test with it and at different ranges to get data from the Wifi signal strength and design the best algorithm to save as much package bursting as it can without losing control. Also, I live in a place in Colombia where average temps are 32 degrees Celsius so that heat in the day shortens battery life. In a few days I'll be going to Bogota where average temp is 18 degrees and that might help get more data with a drastically different temperature readings. So to achieve that I'm currently working in a "black box" functionality to gather as much data as possible for both development purposes and to be used as a flight recorder log for personal analysis. Again, my approach with this app is not just to offer a Windows alternative but to take things in a smooth way to cover the little shortcomings with Tello so for instance both photo downloading and video capturing can be as efficient and effective as possible, which I believe would be the ultimate goal of any app for Tello since it has no SD card. Hope either Ryze or DJI launch a Tello2 addressing those issues.
 

volate!lo

Well-known member
Joined
Dec 13, 2018
Messages
1,725
Reaction score
1,634
Location
Germany
Now its getting interesting. Have you done multiple runs to support these measurements?
Because, just looking at aTello's source code it does not do any package burst at all - yet it's among the worst according to your data.

Have you done these tests using all basic functionality required for a real app? That is do you not only control the drone but also maintain a stable video stream and continuously retrieve telemetry data?

If your measurements turn out to be correct and reproduceable I guess I will have to look into optimizing my protocol handling a bit to achieve better flight times.
 
Joined
Oct 27, 2018
Messages
8
Reaction score
1
Age
36
Location
Cúcuta, Colombia
Now its getting interesting. Have you done multiple runs to support these measurements?
Because, just looking at aTello's source code it does not do any package burst at all - yet it's among the worst according to your data.

Have you done these tests using all basic functionality required for a real app? That is do you not only control the drone but also maintain a stable video stream and continuously retrieve telemetry data?

If your measurements turn out to be correct and reproduceable I guess I will have to look into optimizing my protocol handling a bit to achieve better flight times.
So far I only have scattered data and that's why I'm building a "black box" functionality to gather data as complete as I can and proceed to get data from all the Android apps, even if I can get my hands on an iPhone I could test the iOS app as well. Just then I could publish the results varying several factors that can offer developers a wide view of what's happening and help the others achieve a better functionality.I think I can be doing that by the en of the month after I come back from that trip, so I can have data from different weather, altitude, barometric pressure, etc. For instance, Bogota is much higher than Cucuta, 2600 mts vs 320 mts, and it's also colder, air is thinner and less it's also less windy, but for instance people in the US Canada or Europe can live at lower altitude places where preassure is higher than Bogota but with the same temperature, perhaps more or less wind, so being able to gather data from different places as the drones fly can give me a better perspective on the overall performance of the app, which as you have stated already, ir basically the only thing we can control to use the drone, the Tello itself is dumb and simply obeys the commands, as long they get to the drone, so being able to improve on the fly the way the app works ir crucial to make it fly farther longer and safer.
 

volate!lo

Well-known member
Joined
Dec 13, 2018
Messages
1,725
Reaction score
1,634
Location
Germany
Maybe I don't fully understand what you are trying to achieve because I'm confused why you want to consider temp, pressure, etc. From my understanding temp, pressure, wind, etc are not important for comparing apps and their _relative_ performance. More the contrary is the case: You'd want to standardize external factors as much as possible.

For comparison you have to perform a standardized flight in a standardized setting. Easiest way is to just hover indoors in A/C controlled environment with similar light conditions. It needs to be in a large room and at least 2m altitude so there is little disturbance from prop wash. You need to do multiple runs on the same batteries for each app, and each app has to perform the same basic functions: Accurate drone control, telemetry, video (stable stream including same bitrate, error handling, stabilization, etc). Otherwise its like comparing apples to oranges.

Once you have all this in place I would be very surprised to see any significant difference. 10 seconds or so may be possible but certainly not a minute or two.
 
Joined
Oct 27, 2018
Messages
8
Reaction score
1
Age
36
Location
Cúcuta, Colombia
I want, as you say, compare apple to oranges, but in a logical way. I want to gather as much data as I can from my app and every other I can, from external factors and everything I can. Then I'll sit down and organize all that data to analyze it and build something that can work the best way. Maybe, as you say, some of the data might be useless, but if I can get it and it'll be worth for anything it's not a waste of time. So let's see how things move on and hopefully I'll have some results that again, helps everyone.
 

volate!lo

Well-known member
Joined
Dec 13, 2018
Messages
1,725
Reaction score
1,634
Location
Germany
OK, cool. Let us know when you have results. Maybe I'm wrong and we can really squeeze out a few percent flight time.
 

Rat_rulz

New member
Joined
Jan 1, 2019
Messages
2
Reaction score
0
I want the app on my Xbox so I can control it from my lounge room coach, I think you are onto a winner if it is done right! I see why you have to get very technical to get a smooth stream as if it is not smooth then the experience would be not worth it. So your trying to get it right for market
 

New Threads

Members online

Forum statistics

Threads
3,248
Messages
25,113
Members
8,368
Latest member
Anderson Pires