Firefighters to UK Home Office: Yeah, maybe don't turn off emergency comms network before replacement is ready

As if the UK's emergency services didn't have enough on their plate, the troubled plan to replace their communication network appears to be hitting choppy waters once more. In a pithy document released earlier this month [PDF], Buckinghamshire & Milton Keynes Fire Authority pointed out that the current system, Airwave, is set …

[Stuart Castle]

Business cases?

So, the system was due to go live in 2017, and was heavily revised in 2018, yet the business case to justify it wasn't due until 2020? I would have thought the business case should be done first, as they are usually done so that a company can assess the impact of a project on it's business., and thus decide if it's worth the time and expense.

[Headley_Grange]

Re: Business cases?

I've prepared a number of business cases for a variety of employers and clients. The main skill is the ability to understand/guess the outcome that's required and then to be able to make the business case support it.

There's a good one recently - the A9 dualling cost-benefit analysis showed that the project would return less value than it cost so the Scottish Gov. hired some consultants (not me) to assign a monetary value to the cost of driver frustration. This came to £430 million – £86 million more than the value given to collision reduction - and tipped the project return into the black.

[Michael]

Re: Business cases?

Anyone driving on the A9 regularly would consider dualling the carriage way to be value for money.

[Anonymous Coward]

Re: Business cases?

Just out of interest how much value would you put on the benefits of the road being dualled? If they dualled only on the basis of making it a toll-road to pay for the construction costs how much toll would you be willing to pay?

[Intractable Potsherd]

Re: Business cases?

@AC "If they dualled only on the basis of making it a toll-road to pay for the construction costs how much toll would you be willing to pay?"

Unless there was a free alternative running next to it (say, the existing A9), then £0.00 per mile. If the alternative route did exist, then maybe £0.10 per mile.

[Pascal Monett]

"Anyone surprised at the gigantean cock-up"

. . has not been reading El Reg.

Honestly, telling me about the £3bn overspend is useless. UK Government has pissed away so many billions on so many different failures that one specific figure isn't enough to tell which failure we're talking about.

What is not surprising is firefighters reminding government that hey, that comms system you're going to shut down ? We still need it, and we'll be needing it until you pull your finger out and get the next one working.

[Jellied Eel]

Re: "Anyone surprised at the gigantean cock-up"

Don't worry, it'll be fine.. Capita are on it!

[Psmo]

Re: "Anyone surprised at the gigantean cock-up"

No fair!

[chivo243]

Re: "Anyone surprised at the gigantean cock-up"

Keep it running as long as needed, just don't forget to turn it off when the new system goes live...

[Ashentaine]

Re: "Anyone surprised at the gigantean cock-up"

Or maybe just leave it on standby for a while, as the new one is most likely going to fall over a couple times before they actually get it working properly.

Be sure to also allow for the time spent on rounds of blame assignment in the statehouse and searching the couch cushions to find the budget for repairs, replacement and testing that should have been done before going live, but wasn't for a variety of short-sighted reasons.

[Jimmy2Cows]

Rigorous testing

Don't worry, there'll be no need to run the two systems side by side for a while because ESN will go through the same thorough, rigourous and detailed testing we've come to expect from other multi-billion pound government infrastructure projects.

[Icon: Thusly expected outcome the first time there's a major incident and the comms network buckles under the strain]

[macjules]

Re: "Anyone surprised at the gigantean cock-up"

Acceptance Criteria? If this is 'normal' Capita then 95% SRA is ok. On 1000 emergency calls that is a lot of dead people.

[TRT]

The 5G ESN sponsorship deal went out of the window...

The blue-light sirens were to be re-tuned to emit a sponsorship message...

Huuuuaaaaaweeeeeeiiiiiii..... Huuuuaaaaaweeeeeeiiiiiii.

[Anonymous Coward]

Been saying it for a long while

And it's good that the customers are now saying it as well.

Not only is the mythical next generation not going to be as resilient as tetra (it can't be, the tetra setup was designed with multiple levels of redundancy throughout and that's one reason it cost a little bit), but it's going to be so ridiculously late as to be DOA.

why not wait until 3/4/5G is actually rolled out, and then hop on the bandwagon? Maybe with a supplementary radio for higher data applications, maybe just in vehicles to start with.

Then, once that is usefully working, start considering moving voice to join it. But the tech behind volte doesn't naturally map onto the use case of our emergency services.

And the mobile phone networks can't cope with a simple New Year in London, let alone the required capacity for the London Riots... (in some respects the total traffic should be lower, and they can turn off the system for the plebs, but even so I have serious doubts.

[steamnut]

Re: Been saying it for a long while

I thought that the mobile networks now had some override functions built-in to allow essential services capacity when the normal traffic was blocking the system due to a major event (not a Boy's Own comeback). If not, then it does seem like a good idea.

[Yet Another Anonymous coward]

Re: Been saying it for a long while

It does have a priority user only feature. But it is implemented in the network (obviously) so a million plebs phones try to connect get refused a connection and try again - completely saturating the local airwaves for emergency users that would be allowed to route a call if they could connect.

The feature was a bit of an afterthought to the GSM spec

[tellytart]

Re: Been saying it for a long while

The GSM spec has always implemented this. Because I was working for a broadcaster considered vital infrastructure in 1999/2000, for the period over the millennuim my phone was added to the system as having priority - if I needed to make a call and the local cell had no spare channels, it would drop an existing call to service mine. (The network is also configured to drop any active calls as necessary for anyone trying to make a 999 call so that the closest cell with the best signal will service the 999 call).

[Yet Another Anonymous coward]

Re: Been saying it for a long while

IIRC the problem is that the allow/deny is done by the network behind the masts. So it needs a completed connection to decide if you are priority, it then drops other connections - which retry.

But there is no mechanism for a cell to broadcast that there is an emergency and noone else should try or to tell refused connections not to retry for an hour or for phones to send a priority level in the initial handshake and have the tower decide to continue in the first part of the connection negotiation

[Anna Logg]

Re: Been saying it for a long while

The problems are a bit deeper than that though; an emergency services network (hopefully...) includes end to end redundancy and backup. Providing a meaningful (as in hours) level of RAN power backup in case of mains failure (generally via good old VRLA) is expensive to roll out and maintain. Then there's redundant backhaul etc.

[Anonymous Coward]

Re: Been saying it for a long while

And then there's the testing and maintenance of all of that gear, with regular tests of power cutover, and the occasional emergency cutover that results from something blowing up (but at least that was found on a test, rather than relying on it)

[Yet Another Anonymous coward]

Re: Been saying it for a long while

But the chances of a major powercut in eg central London is 1:1000 and the chance of a major fire incident is 1:1000 so the chance of a fire AND a powercut at the same time is a million to one!

[NetBlackOps]

Re: Been saying it for a long while

Speaking as a field engineer, it's always something you hadn't planned for and if you had, the odds were wrong. Hundred year flood twice in 20 years wrong.

[Jellied Eel]

Re: Been saying it for a long while

Speaking as a designer, I still remember being asked to design network for the ill fated 'Fire Control' network. Which planned to consolidate fire service control centers into a few larger sites. My 'client' was one of the big consultancy firms teamed up to bid on this. So I designed & costed a fully diverse fibre network to each center, complete with east/west building entries for maximum physical seperation. Same with the terminating equipment, with dual racks, CPE and UPS for each rack. Logical network ran as active/active to minimise any failover delays.

Then I presented the solution to the 'client', who said it was too expensive, and could I redo everything using xDSL, but with a higher availability SLA. I explained the technical limitations of xDSL, and why a five nines availability would be.. Unlikely, and the risks of doing it on the cheap for something that was very much a safety-of-life network. 'Client' didn't get it, so I left. That was one of those moments where I swore off doing PFI-type jobs unless I could be sure I was getting my requirements from the end user.

The 'client' in that case wasn't Capita, but such are the joys of PFI.

[Yet Another Anonymous coward]

Re: Been saying it for a long while

It's much cheaper to add another '9' to the SLA - it's literally a key press.

And assuming you are the only bidder on a government contract you know that nobody is going to successfully sue you - the worst you can face is a snide comment in Private-Eye or El'Reg years later

[Jason Bloomberg]

Another delusion crashes and burns

I recall when the loudest argument for making the change was huge cost savings for the tax payer.

I guess I get no applause for having said the cost savings weren't that great anyway, would never materialise, it would be over-budget, late to arrive, not perform as expected, likely never even completed, because any fool could have predicted that with a good degree of certainty.

It's not having done brexit earlier which I blame. Once brexit is done Britannia will again rule the waves and the British Disease will simply disappear. So I'm told.

Most likely we'll just buy Chine^W American. Or keep spunking money up the wall and flushing it down the drain.

[Anonymous Coward]

Re: Another delusion crashes and burns

"I guess I get no applause for having said the cost savings weren't that great anyway, would never materialise, it would be over-budget, late to arrive, not perform as expected, likely never even completed, because any fool could have predicted that with a good degree of certainty."

To be fair, that's just called looking at whatever failed Government IT scheme is in the news this month. :(

[ComputerSays_noAbsolutelyNo]

Why use 2/3/4/5/6/x G at all?

One of the big advantages of plain old airwaves is that it doesn't require a nation covered in base stations. Imagine a large scale power-outage: there aren't sufficient emergency power generators to support (all the base stations of) the mobile network.

I would want my first responders to be able to communicate even when I am no longer able to watch cat videos on the internet.

https://en.wikipedia.org/wiki/Terrestrial_Trunked_Radio

* The much lower frequency used gives longer range, which in turn permits very high levels of geographic coverage with a smaller number of transmitters, thus cutting infrastructure costs.

[chivo243]

Re: Why use 2/3/4/5/6/x G at all?

A colleague is a member of DARES, Dutch Amateur Radio Emergency Service. A quick perusal of their English website shows an UK counter part Ray-net (UK).

https://dares.nl/about-dares/

[Yet Another Anonymous coward]

Re: Why use 2/3/4/5/6/x G at all?

But you can then sell the old frequency spectrum and the new 5G version allows real-time high-definition 3-D video streaming to the scene.

Which is obviously more important than your fire service radio working indoors or underground

[Jellied Eel]

Re: Why use 2/3/4/5/6/x G at all?

Which is obviously more important than your fire service radio working indoors or underground

Hmm.. That got me thinking. So fire burns, coating walls & ceiling in carbon soot. What effect would that have on RF because AFAIK, that doesn't play nicely with carbon.

[hoola]

Re: Why use 2/3/4/5/6/x G at all?

Correct, there is a reason why radio works as well as it does. It is this endless obsession with data and information that has to be provided. So they can have building plans and all that at the touch of a button. These really are only relevant at very large incidents where there will be other organisations involved that already have the information.

In an emergency situation PTT radio is invariably better than some funky phone based device, even if that is trying to emulate PTT.

I have always suspected, that just like FM this is more about selling spectrum to make money than any actual improvements to service on the ground.

Take, DAB, as a package what are the benefits?

Poor streamed quality.

Much is streamed in mono because there is insufficient bandwidth due to cost

Poor reception.

Very bad battery life.

Cost.

Yup, that is a worthwhile upgrade.

[NetBlackOps]

Re: Why use 2/3/4/5/6/x G at all?

All so some Minister, and PM of course, can crow when the spectrum is sold for x number of billions. Ignoring y number of billions wasted to get there.

[Yet Another Anonymous coward]

Re: Why use 2/3/4/5/6/x G at all?

>So they can have building plans and all that at the touch of a button.

So they've been watching too many of those movies where the hero gets to zoom through a 3D wireframe model of all the century old underground tunnels in London? I've had problems getting the floor plans of a building we are currently having built - just to find where I can put some heavy instrument.

Perhaps that also explains the government's love of CCTV - we just put one camera on the top of Big Ben and we can zoom in and "enhance" a crime anywhere in London

[Anonymous Coward]

ESN is never going to provide the coverage of AirWave

Not what I think or say, but the Chief Engineer of one of the mobile companies where I used to work, just as they chucked the tender documents in the bin as it was "too risky" to associate our brand with the system.

[Anonymous Coward]

Uses 5G?

The letter says the new system will use 5G LTE for voice and data. AIUI 5G LTE only exists in a microscopic part of the UK and the chances of it being widespread any time soon is miniscule. More seriously, there are still many places in the UK with no kind of mobile coverage at all. So how is a mobile-based emergency communication system going to work if the emergency is in one of the 'not-spots'? Or indeed pretty much anywhere than a major city centre if it really is dependent on 5G?

[batfink]

I keep saying it...

I'll bet the genius in Motorola who bought Airwave keeps getting a nice bonus every year. What an investment.

[Anonymous Coward]

Remembering Grenfell (and others)

In the first few months post-Grenfell, I seem to recall reports of how well (or otherwise) various aspects of the Emergency Services comms network and core infrastructure and applications performed when it mattered.

I am unable to quickly find any of them today.

Did I dream them?

Pointers welcome.

[alwallgbr]

Re: Remembering Grenfell (and others)

Fireground radios didnt perform well according to reports. However if you look at the allocation plan the problems are self evident.

There is extensive documentation on the enquiry website

[Hubert Thrunge Jr.]

Omnishambles

Whichever way you look at it, it's a shambles.

It was born out of the will to get out of Airwave as fast as possible because the original PFI terms were crippling.

The solution was not ESN, it was hold Airwave to account for the missed performance targets that happened every month, and withhold payment until they sorted themselves out. As Macquarie ran the business with an empty bank account, they would have been forced into administration, and HMG could have bought them for £1 - thus owning the network, and then only having the running costs to pay, not the exorbitant fees that they were being charged. They could then instruct Motorola to upgrade the network to TEDS2 or better, and then go out to tender for a composite device that could do high speed data over LTE leaving critical voice on the now government owned Airwave network.

Nope, can't do that because government seems to be allergic to owning things.

What we must also remember bus that EE won the network deal because everyone else pulled out, and Motorola won the implementation deal because... everyone else pulled out. They could see it was a poison chalice, and would be fraught with issues due to unrealistic timescales with it needing 3GPP standards that still don't all exist yet, and hardware that also didn't exist.

And six months ago they gave Crapita an expensive lifeline to do something that was already easily possible - a bit of VoIP routing. FFS you can do that with a Raspberry Pi and a copy of Asterisk!

Still, it's all keeping me gainfully employed!

I refer the reader to my subtitle.. omnishambles!