I recently suggested in this thread:
http://diydrones.com/forum/topics/need-uav-built-for-under-10k

that Ardupilot would be a low cost way to make a commercial/military UAV.

I haven't used ardupilot in autonomous mode yet, but I've been very impressed with the quality of the ArduIMU.

My impression of Ardupilot is that it is a fully functional autopilot.

I'm curious to get people's opinions: are there any reasons that Ardupilot is not ready for commercial & military use?

If so, what in your opinion are the bare minimum features necessary to a military/commercial drone surveillance aircraft?

Views: 920

Reply to This

Replies to This Discussion

As a technician for the NASA Small Unmanned Aerial Vehicles Lab at Langley Research Center, I can say yes we are using what you call COTS autopilots. Currently we have been using MicroPilot. super expensive equipment but very effective with plenty of features. The reason I have been browsing this site is to gain info and to try out the Ardupilot in one of our airframes to test out its abilities to possibly be a tool we use more often if it fits our needs. My lead Engineer thinks it has what it takes, so we will see what all we can do with it.
Logan

Can you tell just how more advanced, or potentially advanced is MicroPilot compared to the Ardu or Ardu Mega a/pilots. Sure, a lot of time and effort has obviously gone into MicroPilot's GUI, but in terms of just what it is capable of from an aircraft control point of view, just how much more can it do?

With sufficient time put aside, could the Ardupilots be broguht up to much the same aircraft control capability as the MicroPilot.

I was lead to believe last year by a rep at the Farnborough airshow that more than half of the retail cost of the MicroPilot is tied up in its GUI, and that in terms of actual naivigation potential and flight surface control capabilities, the Ardupilot (in particular the Mega version) is capable of pretty much the same performance (?).

Whats your take on that statement?

Thank's

Patrick
Patrick,

Since i have yet to recieve and actually test my Ardupilot Mega I can only comment and compare what Ive read on the Ardupilot to do a fair comparison.

To me, on paper they pretty much are the same with only a few details that im not sure about on the ardupilot. Those being:

Can Ardupilot uplink new waypoint information DURING a flight, changing its course?
- Micropilot can with ease and this is one of our major concerns with using Ardupilot in our vehicles.

Can Ardupilot uplink new gain/speed information DURING a flight?
- Micropilot can and is definetly helpful when you have a flight failure like we had with an aieleron servo going out in midflight. Micropilot accomodated with our uplink of new gains and we flew on one aieleron.

Also micropilots interface is much more user friendly from what ive seen but the trade off is that micropilot is a closed source software program.

These are my questions and when they are answered i can give you a fair comparison on if they are infact just as effective as each other. As soon as I get mine in and integrate it into an airframe and begin testing performance I can give a more detail comparison as well.

Thanks,
Logan Hasenbein

Hi Logan

I'm new to the site and forum. I would be interested to know what the results of your comparison were??? Regarding the issues pertaining to the use of the ArduPilot for military grade projects I too have been debating the issue in my mind. And certainly from a cost/performance/risk perspective with proper analysis it could be justified - provided the required amount of reliability can be achieved - even with commercial grade components  (There is nothing worse than relying on equipment in the field and it does not work). While certainly cheaper, any User would discard a product if found to be unreliable.

However as a concept test bed I think its the ArduPilot (and others) is the "best thing since sliced bread" 

Thanks

Grant

   

Technically speaking ARDUPILOT is DIY kit but produced without any sort of formalised QA testing about its most pronounced functionality, being the autopilot.

In short it is very complicated kit that works, that doesn't fit classification of a commercial rpoduct. It cannot be even exported from US without tons of permissions if it woudl be ready product. It is using commercial components produced in millions.

You will have to spend your time, several months, in order to bring it to work, becaue this is how it is.

Any of some 30 persons actively contributing to its development can screw most recent firmware in many ways, the time before it will be discovered is undefined, and you cannot neither sue nor held ANYBODY responsible for that. So this is not a commercial product as autopilot, but is as fully commercial, working soldered PCB for making amateur autopilots.

If you are asking about milgrade, you must be seriously joking. Milgrade electronics defines specific: design procedures, manufacturing procedures, declaring statistical quality standards, connectors and most of all environmental conditions (temperature range), that practically none of the subcomponents would meet.

Comemrcial electronics defines: specific env. conditions to work, norms like CE or orther electronics compatibility etc.

None were tested because none apply to experimental, one-build apparatus made by amateurs.

Kryz: I think your information is seriously out of date or you have not actually used ArduPilot, since virtually every fact in your comment is wrong. ArduPilot (APM 2) is not a kit. It's sold ready to fly, and is one of the least complicated autopilots on the market. It is not export controlled and can be shipped anywhere in the world (there are nearly 10,000 of them out there). Most people get it running in a hour--it's a one-click software upload, with no programming required. We have a formalized QA process including a full HIL and SIL automated test suite.

I think you need to update your impressions of the open source autopilot world. A lot has changed in recent years. I think you'd be impressed by how comparable they are to closed-source commercial autopilots today. 

I think it's fair to say that even APM2 is not a military or commercial grade autopilot, but that it would make a great basis for a military or commerical civilian UAV.

One of the things you'd have to do to make is so would be to branch the software and lock it down so that it couldn't be open-sourced and you could guarantee a certain level of in-field performance.  Chances are it wouldn't be six-sigma reliability, but you could get it very close for a bunch of non-critical applications in a constrained environmental envelope.

I'll bet there are a bunch of skunkworks in countries around the world that are looking at using APM or something like it for future military applications.

 

"It's sold ready to fly"

How then it is different from ammunition by US export laws? it was the whole point of having a kit...

BTW I am following its design both HW and SW closely just to make sure not a single feature is missing in my design, however only as inspiration, never by code copying.

Commercial grade is defined by terms on which you can sell a product. There are norms for electronic appliances of all kinds.

Kits and experimental equipment are extempt from most requirement about electric compatibility (won't harm, won't shot, wont kill even if getting killed etc). Example, pls read carefully my previous post, CE norms. It if is standalone electronics product, not a component, you need this.

Hi

CE is a simple Self Certification that comfirms ELECTRICAL (not electronic) safety that is not applicable to APM2. It simply tells purchasers that the SELLER has a technical file available for inspection WITHIN europe. it is not relevant in the USA or Canada. It does not certify or prove that the device works!!

 

@David" It does not certify or prove that the device works!!"

And who claims it does or should? You are simply changing topic. Commercial product is commercial product: norms, regulations, some of them fairly realistic and worth investigating.

BTW can you point to a single EU seller that did the procedure some 30KEUR for EVERY make and ardupilot series (since this invalidates with a new HW version), or simply willing blindly to take responsibility? I am just posting an example.

One needs this, another would need ISO9001, others ISOsomethingelse. If you got nothing, wither you are acting on unregulated marked (liiiitle chance) or ignoring something (commercial subcomponent integrators will spot it quick when somebody will ask them for norms).

I am not saying that a product should not be certified, My point was that the CE mark ONLY applies to Europe. The CE mark does not require an expensive certification process as it is a "self certification" process with certain requirements. It is often added to products that are sold outside of Europe but it is not relevant.

Define a commercial product? I can build a new device in my garage and sell it, is it a commercail product, yes because I will pay tax on the sale. Is it  a good commercial product ... possibly,

Is it a military product, NO. But it could be the start of progress towards one.

As a manufacturer for many years and being personably responsible for having many products certified for safety worldwide I will tell you the water is not as defined as you would like it to be. Correct certification is good for everyone but not an absolute requirement in every case. The important words are "What is reasonably practical" in the real world.

 

RSS

Social Networking

Contests

Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.

A list of all T3 contests is here

Groups

Advertisement

© 2013   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service