The number’s up for 999. And 911. And 000. And 111

Opinion It all began on 10th November 1935, when five women burned to death in a house fire in central London. A neighbor had tried to call the fire brigade on his home telephone, but had to wait in a queue for his local exchange. By the time he got through to the operator, it was too late.

Incensed, he wrote a letter to The Times, which provoked an immediate government inquiry. Ah, those were the days. Two years later, the first emergency call system in the world was rolled out: dial 999, a red light would flash at the exchange, and your call would be immediately answered. The rest is history – with many variations like 000 in Australia and 111 in New Zealand, the idea spread around the world and every telephone became a hotline to help.

Except, of course, when it isn't. The UK's 999 system is a regulatory requirement for major telco BT, who recently mucked it up and is in trouble. At least it's a uniform countrywide service, unlike the US 911 system, which is barely a system at all but a patchwork of uncoordinated local provisions. The big problem, however, is technology.

The 999 system set the blueprint, and was brilliantly designed to work with telephones of the day. All phones were powered along copper cables from the exchange, which had battery backup so everything worked through power outages. The system was inherently local and universal: you could pick up a phone anywhere and be in contact with your area police, medics or firefighters in seconds, even if you knew nothing about your location.

Almost none of this is true today. What remains of plain copper cable telephone systems is being phased out in favour of VoIP or mobile, each dependent on an infrastructure that goes out when the light does. You no longer call from a fixed geographic point with a known number, but are at the mercy of a skew of GIS systems that may or may not work for you. If you come across five women burning to death these days, and if you have mobile coverage for your operator, you'll get through but you might not be able to tell them where you are.

This is symptomatic of an even bigger picture, where we know the value of instant reliable emergency communication but the evolution of technology with no coordination has produced a huge tangle of random systems. You may have an emergency network on broadcast radio and TV, but everyone's on Netflix. Or you may use the cell broadcast system: good luck with that. There are innumerable separate systems for aircraft and shipping, many of which are prone to false alarms and none of which provide redundancy for the others.

It's an expensive, unholy shambles and when things really go wrong and a natural disaster knocks out all the local infrastructure you have to wait for the local radio hams to dig themselves out from their cellars and wire things up again. We can do better.

Rethinking the network

Let's go back to 1935 and ask what it is they wanted. A universal system with a robust infrastructure that's extremely simple to use and puts people in immediate contact with those who can help. Digital wireless can provide universal connectivity. You can go anywhere in the world and get connected through cellular or Wi-Fi – or satellite. Satellite is the right answer.

Satellites are fabulous for this. They don't blow over in a hurricane, they don't go out when the local power plant goes bang, and if you can see the sky, you're good to go. As climate change driven weather events make life on the ground ever more precarious, this is a system that's ready to cope.

Aviation and maritime people know this, which is why they use satellites. Very big, expensive satellites with a mess of different systems, to be sure, but that's because that's all there was back when. Not so now.

Make it a condition of giving Elon Musk and Jeff Bezos and whoever that they include emergency universal comms on their satellite constellations to get permission to fly, and we're good. Emergency call services are very low bandwidth, which is very cheap, very power frugal, and very reliable to add to high bandwidth satellites and low cost devices on the ground.

Because we're talking universal standards here, we can bake in solid location detection that we know will work. You can build the user bits into standalone devices or as part of a mobile phone, as long as it's a simple one-button activation.

There is a downside: pocket dialling and kids mucking about are already the bane of emergency responders: fixing this at scale is going to take psychology, UI smarts and very careful filter designing at the call center, but they did all that for 999 in the '30s, we can do it in 2023.

• Microsoft's Azure mishap betrays an industry blind to a big problem
• The FBI as advanced persistent threat – and what to do about it
• The Shakespearian question of our age: To cloud or not to cloud
• The UK's bad encryption law can't withstand global contempt

Yet think of the benefits of a proper universal service. The device can know what languages the user can speak, so the system can route through to appropriate call handlers immediately. It can be adapted for disabilities, special needs or any environmental situation. The backend system, armed with known good location data, can immediately provide the responder, no matter where they are in the world, with a menu of emergency services.

The same system can – and should – work in reverse, carrying vital information about local dangers or precautions. You want the thing to be used; one of the reasons cell broadcast emergency systems don't work is that they're never used enough to be reliable. The old 999 worked because it was based on a universal object in daily use.

As for cost: over time, as existing emergency systems age into expensive obsolescence. The way forward will be to migrate onto this new system, made cheap through scale and reliable through use. One world, one help button: let's think big. The letter to The Times is on its way. ®