Please select Into the mobile phone version | Continue to access the computer ver.
Is Innovation a Curse?
Uploading and Loding Picture ...(0/1)
o(^-^)o
Mike-the-cat
Captain
Flight distance : 21621841 ft
  • >>>
Singapore
Offline

In the fast paced world we live in, designing and building innovative products is generally a positive.

I think it still is, so posters who question the appearance of new DJI products each year should perhaps just smile and be happy with the latest thing they bought and use it for as long as practicable.

Rather than seeing this as 'DJI cheating its loyal customers', I see the breakneck speed of product innovation as a sign that product design and thought leadership in things S&T truly shifting across the Pacific. While the US continues to lead in throwing up creative ideas about how to use gear and ideas for what gear should do, its clear that in terms of engineering and manufacturing prowess, there has been a shift. And that prowess is just ramping up further.

So, call it what you will, but its great that DJI continues to bring to market really great gear. And I hold bragging rights for shorter than I expect, so be it.

DJI - as long as you continue to support your legacy products for at least 4-5 years, you'd be doing OK. Here's where you mustn't forget your customers.


2017-4-23
Use props
WernerD
Captain
Flight distance : 350837 ft
  • >>>
Austria
Offline

Working in the software industry myself, my input would be....
There are two kinds of innovations
  • Evolutional ("Kaizen") where you take a system and optimize it by a little in high frequently. The DJI App would be a good example - adding a minor piece here, fixing a little bug there.
  • Disruptive innovation where you had a great idea which needs a lot of changes and likely different hardware.


Another aspect is How you develop something
  • As a brand new product from ground - no ties hence but it will take time until you are feature complete
  • Continous development - you reuse stuff that exists already and improve. Hence more reliable, faster and less expensive.


The third aspect is a long term vision
  • 20 years ago we would have developed a vision, then an architecture and then started with the implementation. The first 2 parts took multiple years already in order to be perfect. But by then the product idea had been obsolote.
  • Today there is the tendency to focus on a single use case, implement that and then be surprised when reality kicks in and your use case starts to change - you not being prepared for that.
  • Best is obviously a mixture of the two, spend some time on a long term vision in order to be aware of the most obvious things and incorporate them into your architecture (not neccessarily implement them from start but allow for it later).


A good compromise for all the above is Modularity. Let's take DJI for example:

There are the components camera, gimbal, drone body, drone electronics. While most companies separate the four components into four modules, DJI sees camera and gimbal as one unit, drone body & electronics as the second. There are a lot of design interdependencies between camera and gimbal, e.g. masses, center of gravity, dimensions. Makes perfect sense.
As a result, we have two modules and the interface between those two is critical. And that is where DJI is starting to make mistakes.

DJI Ronin-M: A superbe gimbal. Guess what, people want to mount it on rigs like a cablecam, spidecam, crane, drone, tripod. An obvious future use case but the design does not allow for that as the gimbal does neither provide feedback for its current position nor allows for absolute movements ("look 90° to the right"). All it allows is realtive movements - move with a speed of 10°/s to the right and I keep controling the angle manually.

DJI Inspire 2 is the same thing: You have the Inspire 2 with the new cameras at an exorbitant price. And the Inspire 1 is still a good bird. The most logical thing would be to mount a Inspire 2 camera on an inspire 1 bird for a starter. Is there anything that prevents that? Okay, SDCard, SSD and electronics went into the bird instead of being in the camera, so yes. But would it have been a problem to provide e.g. an adapter plate that contains that for the I1? I would think it is doable. And soon people will want to upgrade to the I2 bird for longer flight times, yet with just two components you could serve a much wider audience: I1 with X4s as entry; I1 with X5S; I2 with X5(R); I2 with X5S.
Instead today you have to go all the way to the I2 with its cameras and that is quite a price tag.

DJI Inspire 2 Batteries are another good example: Nobody thought at the I1 development about a dual battery concept. No problem. Experience showed that quite a few I1s fell out of the sky because the battery turned itself off to protect itself. (Okay, now the battery is broken because it fell out of the sky instead of being deep discarded) Can you build a dual battery system with existing batteries? Certainly not, hence you have to change it. Accepted.

But DJI fails even on the most basic task - the maintance of the Go App and fimware - which speaks for itself.

None of that is a problem for DJI at the moment. But it leaves the door wide open for competitors to gain market share. At the moment there are none. Exactly as there were no competitors in the automobile area as nobody can quickly develop a new car and produce it in masses. An attitude that allowed Tesla to pop up.

My point is, all these decisions will hurt DJI longterm. Shortterm it is just a little less revenue - if that - but longterm it prepares the seed for competitors to grow.
2017-4-24
Use props
Mike-the-cat
Captain
Flight distance : 21621841 ft
  • >>>
Singapore
Offline

WernerD Posted at 2017-4-24 03:19
Working in the software industry myself, my input would be....
There are two kinds of innovations

Excellent post and I agree with the points you raise.

DJI seems to be betting their house on the short attention spans of consumers where the bulk of their current revenue comes from. Relative to some other Aussie, Kiwi, North American or European high end offerings, the Inspire 2 is a bargain. However, there is a limit to what price + performance alone will gain you. Higher end customers will not tolerate shoddy service or some 'aesthetic' issues that I think are culturally very differently perceived.

Here is where I see speed bumps appearing as I speculate in this post: http://forum.dji.com/thread-93820-1-1.html

This said, this is a young and smart company. We are part of the theatre of evolution of this product class and its an exciting time to be around.

Would be interested in your response.
2017-4-24
Use props
WernerD
Captain
Flight distance : 350837 ft
  • >>>
Austria
Offline

DJI Goggles is another obvious example.
If I read correctly, they support the DJI Mavic directly, only.
Inspire 2 is supported via USB - potentially more latency hence and the stupid cable.
But imagine I put a Ronin on a crane and can follow the object by moving the head. Would be nice, wouldn't it?

Too little thoughts into modularity in my opinion.
2017-4-26
Use props
Mike-the-cat
Captain
Flight distance : 21621841 ft
  • >>>
Singapore
Offline

WernerD Posted at 2017-4-26 00:07
DJI Goggles is another obvious example.
If I read correctly, they support the DJI Mavic directly, only.
Inspire 2 is supported via USB - potentially more latency hence and the stupid cable.

Mavic is supported via mini-USB and the latency is 110ms
Inspire 2 is supported via HDMI and lacks many of the navigation features of the Mavic (which has the richest feature set of the craft supported). The latency is 140ms and well within reasonable... see https://dl.djicdn.com/downloads/ ... _User_Manual_EN.pdf

While the design decisions seem odd, they appear deliberate.
My guess is that DJi anticipates that professionals would prefer to use the Cedence / Crystal Sky combination for work and just plug in a Goggle for fun.
As Mavic doesn't give you an option to this combo (makes no sense anyway to have a RC controller+ display twice as heavy as the aircraft!!!), they loaded up the features on Goggles to make it more fun.

If you are going to shoot, by all means, but do check public domain available information before you do so

I'm not complaining anyway. High Res 1080p standard goggles for $480? No wonder Zeiss stopped making the Cinemizer some time ago...
2017-4-26
Use props
WernerD
Captain
Flight distance : 350837 ft
  • >>>
Austria
Offline

Hmm. I quote from the linked manual:

MavicPro: The goggles can be connected wirelessly
Inspire 2: Connect via USB. So you have hold the remote in your hands and/or use a long usb cable. Latency I don't know the effect of 0.1 secs. Seems to be high, but that's beside the point anyhow, as the gimbal will respond much slower. Is motion sickness hence a problem? Don't know.
Inspire 1: HDMI only, no head tracking. A pure image transmission and for that you are better off with a tablet, I guess.

But the main point is, Ronin: No.

So at least the latest Ronin, Ronin MX (Lightbridge enabled) and the I2 should have been included in the same quality as the Mavic. That's my point.

Better?

Let's bet: There will be an Inspire upgrade enabling the goggles fully, within 12 months. An Inpire 3 with side cameras and LIDAR ground sensor instead of the ultrasonic ones (Airdog has that) maybe?


2017-4-26
Use props
Mike-the-cat
Captain
Flight distance : 21621841 ft
  • >>>
Singapore
Offline

WernerD Posted at 2017-4-26 01:11
Hmm. I quote from the linked manual:

MavicPro: The goggles can be connected wirelessly

My bad about the Mavic - it is connected wirelessly but the Inspire info is correct
For INSPIRE series and Phantom 4 Pro+/Advanced+ aircraft, connect the HDMI port of the Goggles and the remote controller with the HDMI cable. Please note that this will be live view only.               
10ms = 0.01s not 0.1s
The latencies quoted in the manual are superior to those one can get connecting an iPad Air2 directly to the current controller

Psst - don't consult for free! Get paid for it




2017-4-26
Use props
WernerD
Captain
Flight distance : 350837 ft
  • >>>
Austria
Offline

Aaargh... was  typing too fast and missed a zero.

I love those strategy mind games. But getting paid for it would mean these are insights nobody else has. Don't think this is the case, these are obvious organizational problems and I am sure DJI knows about those. Usually there are other reasons why you can't execute the needed changes.
2017-4-26
Use props
RichJ53
Captain
Flight distance : 1837356 ft
  • >>>
United States
Offline

WernerD Posted at 2017-4-24 03:19
Working in the software industry myself, my input would be....
There are two kinds of innovations

Thank you for taking the time to write this excellent summary. I couldn't agree more

All the best
Rich
2017-4-26
Use props
Advanced
You need to log in before you can reply Login | Register now

Credit Rules