Results of the AIS Consultation meeting Tue 8 August 2006

[Mike Cross]

With my AOPA hat on I attended a customer consultation meeting at Heathrow on Tuesday 8 August.

You can find a synopsis of the meeting here. (http://www.joinaopa.com/forum/viewtopic.php?p=417#417)

If you have any questions I'll be happy to respond.

Mike

[dublinpilot]

Well done Mike. Sounds like a useful meeting.

Phil would like to move to an independent Web interface, over which UK AIS have much more control. He would like to be able to do this inside a year but recognises that this is probably not achievable.


A year to make a website? I'm not a web developer, but a year does seem like a very long time to me. Drauk seems to be making a notam website in his spare time. Surely a year is way too long?

Any of you IT guys care to comment?

Or have I got the scope of what's envisaged wrong?

dp

[rustle]

Any of you IT guys care to comment?

Or have I got the scope of what's envisaged wrong?

dp

I don't think it is the scope that's the problem...

Have you ever tried to get anything approved by the CAA? ;)

12 months is optimistic! :O

[Fuji Abound]

Mike

You asked for suggestion on this forum and I felt some useful suggestion were made. It doesnt seem to me these were discussed. I wonder why not?

There were numerous comments made by many on this forum as to how unuser friendly they found the AIS web site - this does not seem to have been conveyed.

"There is a lack of appreciation by users of all of the issues involved in NOTAM production and dissemination" - if indeed this really is an issue the AIS carry no information about the process on their web site in order to educate those who might want educating. In the alternative, it seems an all too ready excuse, when maybe the average pilot is not remotely interested in either the production or dissemination - he simply wants a relaible and user friendly means of accessing vital flight safety information.

[Mike Cross]

Rustle & DP

I don't think either of you have got it particularly right.

The fact that Thales IS have been having problems indicates that it's not particularly simple. The dynamic data is changing constantly and the "static" data is subject to change through the AIRAC cycle. The whole thing has to be fully compliant with ICAO SARPS and the process has to be entirely auditable with an audit trail for all transactions and be able to serve up an average of over 1000 individual briefings a day, derived from NOTAM from nearly 100 different worldwide NOTAM offices.

In the normal course of events you would think that fixing something that is already working up to a point would be, relatively speaking, easier than starting from scratch, however that is proving not to be the case.

The core database, input/output routines and the multi-user UNIX environment in which it works is already there and the design authority for that rests with others so you can't go fiddling with it. There are complex communications feeds as well.

The people who have expertise in the data do not have expertise in the computing skills involved, and the people with computer skills who will end up doing the job have no knowledge as yet of the data structure, the requirement, or the constraints imposed on the data by international treaty.

The users range from the Airlines, Military, ATSU's, AOC Operators, Glider Pilots, Hang Gliders, Microlighters, Parachutists, Old Uncle Tom Cobbley and All, right down to the recreational pilot and the sheer bloody-minded stick in the mud who can't be @rsed to learn anything.

The professionals who were at the meeting and who take the data feed from the core database don't find it so simple that they can knock off a website in their spare time that the esteemed members of this august forum will fight to pay for.

There is no money in it from users and the budget has to be fought for and the project defined before anyone sits down and tries to code it. If you think it's easy I'd love to see you try:E

They did suggest I might like to design it for them. I may be stupid, but not THAT stupid.

Get a copy of ICAO Document 8126 (even that's not particularly easy) and see how you get on.

FA
Be assured the concerns are being addressed. I put together two papers based on the points raised on this and other forums. One was to do with the design and user friendliness of the site and the other was to do with the content. Both were sent to AIS in advance of the meeting and passed round the appropriate people. Some of the content issues were raised, as indicated, and are going to be discussed at the next interface meeting between AIS and AUS.

The website design issues were not of interest to British Airways, Bytron, or AvBrief because none of them use it so I did not wish to hog the time on something that was not of interest to the majority of participants. AIS are working on the suggestions made and I expect to see some changes happening over the next few months. Sorry if I did not make this clear.

Mike

[egbt]

Mike,
I wish Thales were doing their (paid) job as well as you are doing your (unpaid) one.
I’m sure this is not as easy as it appears from the outside but if those of us in industry were delivering such poor availability on such a (safety / business) critical system someone would have been in deep do-da long ago. And by now I, as the head of IT for my organisation, would be getting concerned about my own position.

[Fuji Abound]

Mike

Many thanks for the update, response to my post and all your work.

[FlyingForFun]

Is there any possibility that AIS (or whoever) might invite formal tenders for a system to take care of this requirement ? After all, there is obviously wide-spread unrest amongst the user base ... if not, why not ?

If so, where do I (and I'm sure, many others) apply ?
I wonder how difficult it would be to get together a group of PPRuNers with IT experience, form a limited company, and then have that company bid for the project?

In other IT environments where I've worked, a substantial number of the professionals involved in development have had an extensive knowledge of the business (sometimes because they have worked in one specific industry for many many years, other times because the system is public-facing, so they can relate to the end users). I wonder how many of Thales development team are pilots, or have any inside knowledge of the aviation industry? Surely a group of "insiders" would be ideally places to put together something suitable?

Unfortunately, web development was never something I got heavily involved in before I left the IT industry - but I could certainly offer my services as a project manager, compile specifications, specify user interfaces, and develop any back-end systems.

FFF
--------------

[rustle]

Rustle & DP

I don't think either of you have got it particularly right.

The fact that Thales IS have been having problems indicates that it's not particularly simple. The dynamic data is changing constantly and the "static" data is subject to change through the AIRAC cycle. The whole thing has to be fully compliant with ICAO SARPS and the process has to be entirely auditable with an audit trail for all transactions and be able to serve up an average of over 1000 individual briefings a day, derived from NOTAM from nearly 100 different worldwide NOTAM offices.

Mike

I don't buy that. Sorry.

The fact that Thales IS have been having problems indicates that they are not putting sufficient effort into fixing it.

It is their front end, they wrote it, they support it, they should have fixed it by now.

The database behind the UK version cannot be a million miles different from the one behind the Austrocontrol site (which Thales also implemented and which works), nor can the data types, AIRAC cycles, ICAO SARPS be that different in another EU country.

The excuses for not having it fixed are almost as pathetic as the "fixes" implemented so far. :yuk:

[OVC002]

There does seem to be an element of "re-inventing the wheel" here. Thales (a French company) could possibly have a word with the DGAC. Their Olivia system seems to work, is easy to use, and is free of charge:eek:

[Mike Cross]

Please don't take me as a defender of Thales, which I'm not.

However neither am I a supporter of the school of thought that says it's a piece of cake and anyone can knock it off in their back bedroom in a couple of evenings.

Those who want a go form an orderly queue and write to

Phil Bate
Manager, UK AIS
Control Tower Building
London Heathrow Airport
Hounslow
Middlesex
TW6 1JJ

asking him to keep your letter on file and to contact you when things start to happen.

FFF
Thales are experienced in the Aviation Industry, the main back-end system works well and AIS are very satisfied with it. It's the web interface, which is a separate application, that is the cause of the ructions. Their engineers seem to have similar experience to your own.
Unfortunately, web development was never something I got heavily involved in before I left the IT industry - but I could certainly offer my services as a project manager, compile specifications, specify user interfaces, and develop any back-end systems.

Mike

[rustle]

However neither am I a supporter of the school of thought that says it's a piece of cake and anyone can knock it off in their back bedroom in a couple of evenings.

Mike

Neither am I ;)

However this error 500 has been ongoing since October 2005. :yuk:

See HERE (http://forums.flyer.co.uk/viewtopic.php?p=335449#335449) and follow the link therein.

Hardly a couple of evenings ago. :p

PS, Phil has my details already ;)

[IO540]

be able to serve up an average of over 1000 individual briefings a day

I don't want to make a cheap snipe here but you could easily serve 1000 briefings a day over a 256kbit/sec home ADSL uplink :)

It's just text, 10-20k bytes per report. You could serve dozens of times more data over a £25/month uncapped ADSL line.

This also can't be a server loading issue - unless they are running this on an old laptop, which I am sure they aren't. Maybe it's written in Java?

Something else is going on.

[Saab Dastard]

I have to join the growing band of Thales-knockers on this.

I was involved in the design and delivery of a web-front-end for a UK rail ticketing application, which I believe was at least as complex as the AIS back-end, and this was initially delivered in a lot less than a year.

Even the first release was handling way more than 1000 transactions a day! (and it was actually over a 256K link, IO540!)

There must be something more than purely technical issues behind the delays.

SD