Opinions sought on User Interfaces for Machine Control Applications.
- From: "Paul E. Bennett" <peb@xxxxxxxxxxxxxxxxxx>
- Date: Fri, 01 Dec 2006 20:29:53 +0000
Regardless of what machinery is being controlled and what the user's view of
the system looks like on the desk (whether it is a web-based system or a
control mimic) I am seeking some opinions from the rest about what the most
appropriate response time maxima should be. I have found, followed and read
a few links and so far the best advice gives three zones (0.1 seconds, 0.2
to 1 second and 1.1 seconds to 10 seconds) with advice on the style
expected for each of these zones.
<http://www.useit.com/papers/responsetime.html>
<http://www.developer.com/design/article.php/1545991>
<http://hcibib.org/sam/>
<http://blogs.ittoolbox.com/eai/implementation/archives/human-interface-evaluation-12534>
<http://www.sigchi.org/bulletin/1996.2/Chris-Johnson.html>
<http://www.bnl.gov/humanfactors/0700_Rev2.asp>
What I am looking at would be some quotable figures that would apply to an
embedded (perhaps PLC based) system being conrtrolled through a centralised
server based farm of terminals.
Some parts of this conglomorate system gives responses in the two seconds
region (by being given a little more priority for the sub-system traffic).
Other parts are operating with a latency, between commanding a plant action
and seeing confirmation of the action having occurred, of four to 8
seconds. This does not feel so good to operate.
In operating with the programmers interface to a PLC I was able to confirm
that toggling the bits that caused the actions to occur and seeing the
response status from the plant was seen within 3 scans (28ms per scan).
This was for the operation of a small, fast, solenoid valve. With the
central user terminal system this same action was at least 4 seconds and
almost took 8 seconds on occasions (first scan sets the output, one scan
while the device changes state physically, the third scan returns the new
state of plant - approximately 84ms end to end of the change of state).
This indicates to me that the central system is much too loaded and could do
with a boost in capability or an off-loading of some of the work that it is
carrying out.
So, how swift should I specify the plant change of state response be
displayed on the users terminal? This figure would be more interesting if
it was also backed by some good document links. In the mean-time I am going
to see if I can write a justification to support demanding less than 0.25
second response to any change of state of the plant.
--
********************************************************************
Paul E. Bennett ....................<email://peb@xxxxxxxxxxxxxxxxxx>
Forth based HIDECS Consultancy .....<http://www.amleth.demon.co.uk/>
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************
.
- Follow-Ups:
- Prev by Date: Re: Embeddded programming, define embedded.
- Next by Date: Re: Opinions sought on User Interfaces for Machine Control Applications.
- Previous by thread: Heap
- Next by thread: Re: Opinions sought on User Interfaces for Machine Control Applications.
- Index(es):
Relevant Pages
|
