GPS loss plus faulty inertial code caused fatal Camcopter crash

From New Scientist (via SAUSNews), a report on the fatal UAV accident in South Korea last week:

In a statement supplied to New Scientist, Schiebel says its Camcopter S-100 drone, a 150-kilogram rotorcraft capable of 220 km/h flight, should have coped in any case because GPS can be lost for many reasons, such as an inability to access the positioning satellites due to obstruction by high buildings. The Camcopter has multiple inertial measurement units that “allow safe operation and recovery in the absence of GPS signals” the firm says.

“All information recovered to date indicates that after a loss of GPS signals to the aircraft’s receivers incorrect handling and omissions over a time period of a number of minutes, resulted in an unfortunate chain of events that ultimately led to the crash,” the statement says. Emergency procedures “to ensure a safe recovery in such a situation” do not appear to have been “correctly and adequately followed” it alleges.

The Schiebel aviation engineer who died – a 50-year-old Slovakian with much experience of the technology – was assisting two remote pilots working for one of Schiebel’s South Korean partners. He was not in control of the aircraft, the firm says. It’s thought the Camcopter was being tested for new duties in border operations.

The accident aircraft had been used by the South Korean authorities since 2008 to police major events – such as the 2010 G20 summit in Seoul, says a Schiebel spokesperson.

Views: 4416

Comment by Alex Arevalo on May 18, 2012 at 12:51pm

So, I guess there was a software bug?

Comment by Jack Crossfire on May 18, 2012 at 2:10pm

There's nothing about faulty inertial code, unless Wired has a mole.  N Korea jammed their GPS, a human took over, & as usual for a human, he crashed it. 

Comment by Alex Arevalo on May 18, 2012 at 2:23pm

From the first sentence in the statement:  "Schiebel says its Camcopter S-100 drone...should have coped in any case because GPS can be lost for many reasons,...".  

That implies that there was code to handle the lost of GPS signal. It still sounds like the software failed or it maybe that function was never tested. Many military systems have to have Built-In-Test (BIT). That BIT should have been able to turn off the GPS and check for the presence of other backup navigation/control signals as part of the testing done at start up.

I suppose that the jamming of the GPS signal could have jammed other signals, in which case the software could have run out of backup options.

Comment by Jack Crossfire on May 18, 2012 at 5:34pm

The fact is the loss of GPS allowed an alien tractor beam emanating from an underground spaceship which was buried in the last ice age to take over & pull it towards the ground.

Comment by Alex Arevalo on May 18, 2012 at 6:07pm

Oh! Is that what happened? I hate it when that happens! Who woke those guys up, anyway? Must be because of global warming. They finally thawed out. Still it can't have been any fun to land your UAV on your own vehicle!

Comment by Carl La France on May 18, 2012 at 6:45pm

I was  telling Gary Mortimer last week  I remember many times sitting stationary  on the tarmac in different big planes with the electronics running by the buildings  the VOR DME  and the Loran would be right on  after awhile the GPS - DME although it was still working and receiving a signal would  start" hunting"trying to find it's self? floating all over the place showing us 50 to 150 feet from  our position . As soon as we started to move the GPS would straighten up and work fine. GPS doesn't show you where you are but where you have been . You have to be moving for it to work effectively . You move it updates,you are always "just ahead " of your position  You remain stationary long enough it will start to wander . If those guys had the drone hoovering stationary 20 feet away outside the van near buildings relying on GPS for position any length of time   it is a recipe for disaster .   Even with a alternate source of guidance  they were manovering to close it would take just a puff of wind. The van looks like it got hit from something out of "Robo Cop" Sad some one died .They need to extend their "Buffer Zone "


Developer
Comment by John Arne Birkeland on May 18, 2012 at 7:00pm

Emergency procedures “to ensure a safe recovery in such a situation” do not appear to have been “correctly and adequately followed” it alleges.

a 50-year-old Slovakian with much experience of the technology – was assisting two remote pilots working for one of Schiebel’s South Korean partners. He was not in control of the aircraft, the firm says.

This is just a polite way of saying that a human messed up, and sadly they payed the ultimate price.

Comment by Ellison Chan on May 18, 2012 at 8:05pm

I say the safest thing that a craft capable of hovering should do in case of GPS loss is to do just that, HOVER.   No RTL/RTH, not even landing, since you don't know what's under the craft at the time.  The landing should be done under IMU stabilization, under manual pilot control, whenever it is deemed safe.

Comment by Robert Sinclair on May 18, 2012 at 9:35pm

Ellison, Agreed.  Taking from what John wrote and what was written in the article maybe a remote pilot was trying to RTL.  I wonder if we'll ever get more data from this tragedy?


Moderator
Comment by Gary Mortimer on May 18, 2012 at 11:53pm

There is more to this, Paul and I are hard on the trail.

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

© 2017   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service