Arnav crashes

I just finished my SR22 training with Wing Aloft. On a flight yesterday, I was exploring the Arnav options and found a sequence of button pushes which crashed the unit in the plane I was flying, N140CD, each time I tried it.

Here is the sequence:

go to menu page

Slew

Ident

Sel

Ack (cursor moves down one line to next field)

Sel (cursor moves over one space)

Next

The display then mostly goes blank, but does flash repeatedly, until rebooted, a DOS screen which shows the DOS command which likely launches the whole Arnav operating program.

I’m not so worried about this bug per se, but I do worry that Arnav has perhaps not tested their software as well as they might. If this odd sequence crashes the Arnav, perhaps some other situation, which is more likely in flight, could happen. And if they haven’t tested their software well, how robust is the hardware?

Not a huge deal, but it would be interesting to know if this bug existed in other planes (N140CD is a demo model with about 800 hours on it). I would also like to know what Arnav’s quality control process is for software development (fwiw, I manage three groups at a company that builds both chips and complex software and two of these groups do nothing but build software and hardware to test hardware and software).

I know that the Arnav is for “VFR only, monitoring purposes only”, but the fact is it is quite useful in the clouds and if it crashes on you at the the wrong time, that could be quite a bummer.

Other miscellaneous notes:

  1. I like the Arnav! And I’m getting the engine monitoring package installed.

  2. I found the factory tour facinating. They really seem to be in high gear.

  3. I have several reports of a/c delivered with minor squawks which the factory should have been able to find before the customer got it. Make sure to turn the heater on for a while to make sure it works and is not blowing fumes (which heaters tend to do the first time they are used).

  4. New owners are naturally excited by seeing their new a/c and respond to the warm hospitality of the Cirrus folks and Minnesoteans in general, but rememmber that Cirrus isn’t perfect and they are straining to meet production goals. It is much easier to have problems fixed at the factory before you leave.

  5. Cirrus production has done a poor job in the last few weeks (and in the past) of predicting when a plane would be ready. They underestimate. When you get a date, even three weeks before delivery, you might want to add a week or two to be sure you won’t be having to change all your plans at the last minute.

  6. In N140CD I measured 100 Dba in the front seats, around head level, and 95 in the back.

  7. Despite what Gami said about tuned induction engines having imblances, I found the engine in the plane I flew to be amazingly well balanced as evidenced by being able to lean way past peak without roughness (I didn’t leave it there, of course).

We used to have similar problems in military aircraft, even with rigorous software specs. In the P-3 I remember that imaginative crewmembers could come up with some really strange sequences of button pushes that would cause problems. Lockheed’s answer usually was: Frankly, we didn’t think anyone would do that! Actually, not a bad reply. It is, literally, impossible to create complex software that is completely bug-proof. Even improving reliability from say 99% to 99.5% might well cost millions of dollars and result in months or years of delay.

I’m not familiar with the Arnav, but is the button sequence that caused the problems one that a pilot would normally be expected to use?

Joe

I just finished my SR22 training with Wing Aloft. On a flight yesterday, I was exploring the Arnav options and found a sequence of button pushes which crashed the unit in the plane I was flying, N140CD, each time I tried it.

Here is the sequence:

go to menu page

Slew

Ident

Sel

Ack (cursor moves down one line to next field)

Sel (cursor moves over one space)

Next

The display then mostly goes blank, but does flash repeatedly, until rebooted, a DOS screen which shows the DOS command which likely launches the whole Arnav operating program.

I’m not so worried about this bug per se, but I do worry that Arnav has perhaps not tested their software as well as they might. If this odd sequence crashes the Arnav, perhaps some other situation, which is more likely in flight, could happen. And if they haven’t tested their software well, how robust is the hardware?

Not a huge deal, but it would be interesting to know if this bug existed in other planes (N140CD is a demo model with about 800 hours on it). I would also like to know what Arnav’s quality control process is for software development (fwiw, I manage three groups at a company that builds both chips and complex software and two of these groups do nothing but build software and hardware to test hardware and software).

I know that the Arnav is for “VFR only, monitoring purposes only”, but the fact is it is quite useful in the clouds and if it crashes on you at the the wrong time, that could be quite a bummer.

Other miscellaneous notes:

  1. I like the Arnav! And I’m getting the engine monitoring package installed.
  1. I found the factory tour facinating. They really seem to be in high gear.
  1. I have several reports of a/c delivered with minor squawks which the factory should have been able to find before the customer got it. Make sure to turn the heater on for a while to make sure it works and is not blowing fumes (which heaters tend to do the first time they are used).
  1. New owners are naturally excited by seeing their new a/c and respond to the warm hospitality of the Cirrus folks and Minnesoteans in general, but rememmber that Cirrus isn’t perfect and they are straining to meet production goals. It is much easier to have problems fixed at the factory before you leave.
  1. Cirrus production has done a poor job in the last few weeks (and in the past) of predicting when a plane would be ready. They underestimate. When you get a date, even three weeks before delivery, you might want to add a week or two to be sure you won’t be having to change all your plans at the last minute.
  1. In N140CD I measured 100 Dba in the front seats, around head level, and 95 in the back.
  1. Despite what Gami said about tuned induction engines having imblances, I found the engine in the plane I flew to be amazingly well balanced as evidenced by being able to lean way past peak without roughness (I didn’t leave it there, of course).