W
Winfield Hill
- Jan 1, 1970
- 0
-
Categories
-
Platforms
-
Content
data-stream clock with just a little doppler shift
J
John Miles
- Jan 1, 1970
- 0
Another good article by the same guy:
http://msnbc.msn.com/id/6301146/
"After all, these switches were reportedly developed as a nuclear
warhead safety device, so one could just assume that they were properly
wired." Um... did the nuke guys make the same assumption?
Amazing how it's usually the simple stuff that hoses the mission.
Someone mounts the G-sensors upside down (Genesis), conflates metric and
English units (Mars Climate Orbiter), misses a bug in some mundane user-
interface code (Therac-25), or fails to check the specs of reused
software components against the requirements of new hardware (Ariane 5,
and now Huygens).
How do we treat human error as an input to our otherwise-exhaustive
engineering models? That question seems to be worth its own
disciplinary field.
-- jm
P
Paul Hovnanian P.E.
- Jan 1, 1970
- 0
John said:
Randomly select a sample of manufactured units and test them.
Good question. I wonder to what extent each of the aforementioned goofs
can be traced back to a requirements or design communications error
between the systems level engineering folks and the component or
subassembly design/manufacturing groups.
If you model each component of the design and manufacturing team as a
node on a graph and the data flow between each as a link, an error rate
and latency can be assigned to each link depending on such things as its
length and method (the people in the next cubicle vs. an engineering
group a few time zones away) and the cost each node must incur to
correct errors in the communications channel or time wasted working with
outdated data.
J
Joerg
- Jan 1, 1970
- 0
Hi Paul,
and, consequently, failure to adhere to a proper review process. Quote
from the article:
Quote: "JPL's Horttor admitted that NASA probably could have insisted on
seeing the design if it had agreed to sign standard nondisclosure
agreements, but NASA didn't consider the effort worthwhile,
automatically assuming Alenia Spazio would compensate for the changing
data rate." End of quote.
We see that a lot these days, unfortunately. Especially in the world of
software. Either one party assumes the other party did it all just fine,
or one party decides that a certain information does not need to be
disclosed.
Regards, Joerg
In this case it seems to boil down to a serious lack of communication
and, consequently, failure to adhere to a proper review process. Quote
from the article:
Quote: "JPL's Horttor admitted that NASA probably could have insisted on
seeing the design if it had agreed to sign standard nondisclosure
agreements, but NASA didn't consider the effort worthwhile,
automatically assuming Alenia Spazio would compensate for the changing
data rate." End of quote.
We see that a lot these days, unfortunately. Especially in the world of
software. Either one party assumes the other party did it all just fine,
or one party decides that a certain information does not need to be
disclosed.
Regards, Joerg
R
Rich Grise
- Jan 1, 1970
- 0
Ah, what a fantasy! George the Anointed pushes the button, the missiles
all take off, fly their majestic ballistic trajectories, and ten
thousand multiple independently retargetable reentry vehicles go "plop."
My heart wants to sing just thinking about it!
Cheers!
Rich
P
Paul Hovnanian P.E.
- Jan 1, 1970
- 0
Joerg said:
Right. Now multiply this by the number of inter-organizational
interfaces that exist within a project. Granted, these interfaces also
exist between groups within an existing organization. But when the
decision is made to cross corporate boundaries, what was shared without
concern for cost or information security must now be covered by
contract.
J
Joerg
- Jan 1, 1970
- 0
Hi Paul,
contract and a good dose of mutual trust. We have had a lot of
cooperation with outside companies and other organizations in the
medical field. So far there was never any breach of confidentiality, no
IP scuffles and we were totally open with each other. In medical you
just have to open in order to achieve an efficient and most of all safe
design. Everything needs to come together on the table during a design
review, no hidden code, no hidden HW architectures.
Regards, Joerg
Yes, it has to be covered by contract. But this can be one single
contract and a good dose of mutual trust. We have had a lot of
cooperation with outside companies and other organizations in the
medical field. So far there was never any breach of confidentiality, no
IP scuffles and we were totally open with each other. In medical you
just have to open in order to achieve an efficient and most of all safe
design. Everything needs to come together on the table during a design
review, no hidden code, no hidden HW architectures.
Regards, Joerg
D
dd
- Jan 1, 1970
- 0
I believe cellular phone base station receivers have the sameWinfield Hill said:
characteristic introduced deliberately.
Max velocity for mobile 280km/sec.
A correspondent in the Science mag New Scientist gave the wrong answer
recently on their back page questions answered as the RF bw and not the
modulation rate was employed to calculate doppler tolerance.
Gave a result orders of magnitude out.
I emailed to point out their error without any response, and so it
goes.
dd
