All tech must connect

When you receive an e-mail, text message, video or MP3, are you really bothered about how you got it? REAN VAN NIEKERK, MD of Metacom believes many users don’t really care how their data is delivered. Their main concern is that it arrived on time and that the data is intact and has not been tampered with.

So long as it arrives on time, do you care whether your email is delivered to you via WiFi, ADSL, 3G, GPRS, satellite, WiMax, Diginet line, high-frequency radio or tin cans and string? No? We thought not.

More and more, the actual technology which is used to transmit and deliver information is irrelevant. Provided the communication arrives intact, on time and unmolested, the job is done. At Metacom, we believe that means building not just networks, but integrated communications infrastructures that can deliver any message, anywhere, reliably and securely. Whether you’re in the middle of Sandton or the middle of the Kalahari should make no difference.

This is particularly important for the kinds of machine-to-machine industrial communication that are increasingly running all kinds of complex systems, from the electricity grid to the sewage network to vehicle fleets. It’s not just people who communicate anymore. The technological world we have created is alive with billions of sensors and recorders, all busily measuring, monitoring and reporting back on their tiny area of responsibility.

Eskom’s distribution automation is a good example of how this works in practice. Power lines criss-cross the country, delivering electricity to the largest cities and remotest farms ‚ and every now and then there is a fault on one of those lines that can result in power outages over large areas. In the old days, finding and fixing a line fault could take hours. Nowadays, sensors all over the network feed information to intelligent controllers at major substations. When a portion of the line goes down, the fault is not only found and fixed more quickly ‚ the controller also isolates the fault so much larger sections of the network can get quickly back online in the meantime.

There are many similar examples: traffic lights, water pumping stations, weather stations, vehicle trackers, assembly lines, ATMs and credit card machines all communicate nowadays.

Often, this communication is very important. Losing an email about a cute cat picture is no big deal: missing a fault alert from a critical piece of infrastructure is a very big deal indeed. A communications infrastructure for machines needs to provide reliability, security and speed ‚ and it needs to do so as cost-effectively as possible.

Any integrated communications infrastructure must get several things right if it is to succeed in this task. First, it mustn’t discriminate between technologies ‚ it should be able to use anything that is available, from satellites to 3G to carrier pigeons if necessary, to deliver its message. Secondly, like the internet, it must be system fault tolerant. In other words, if there’s a problem in one part of the system there must be alternative routes for messages to take. Taking both of these conditions together results in something that is far more reliable and robust than any individual communications network. We believe a well-designed communications infrastructure should be able to approach 100% reliability.

System fault tolerance also suggests the third aspect of a communications infrastructure: it must contain enough intelligence to be able not only to detect, report and where possible repair faults, but also to make complex routing decisions. This is where artificial intelligence is going ‚ not towards the ability to mimic human beings, but towards the development of highly specialised algorithms and processes for handling fast, complex but ultimately rule-bound decision making.

Finally, a successful communications infrastructure must make its edge ‚ the place where it interacts with clients and the rest of the world ‚ both easy to access and highly secure. This means what at Metacom we like to call ‚edge enablers‚ ‚ devices that offer reliable, secure, uncomplicated access to the greater infrastructure. Our goal is for our clients never to have to think about what goes on between edge enablers: If you have a device in Johannesburg and another in Cape Town, they should both appear on your network as if they’re in the same room. What happens in between doesn’t matter.