21.09.16


Connectivity – often the least considered element of IoT


You wouldn’t start to build a car by buying the paint first. Yet, this is how many organisations are approaching their IoT projects. The tendency is to focus on the sexy bits that people can see, with little consideration to the ‘nuts and bolts’ stuff that will make everything work -- and the most fundamental of these is the connectivity. This may seem like a ‘boring’ commodity, but it should not be the last thing people consider. Connectivity is the critical enabler, and it’s high time it was taken more seriously.

Many companies approach us when they are somewhere between the genesis of an idea and the point where they are about to go to market. Too often it’s towards the end of this timeline. Although the ideas are often brilliant – it might be a UBI telematics dashboard, a healthcare app or a tracking device – all too frequently the project team has left one of the most fundamental elements til last.

It may sound daft to say, “consider the connectivity” when it comes to IoT project but you would be surprised how frequently this element of the technology is taken for granted, causing people to hit obstacles later down the line.

Reliability

Connectivity needs to work more than most of the time in the right circumstances; an IoT device has to be reliable. It’s no use finding out that a flood sensor placed on an upland watercourse has great signal in winter but only intermittent connectivity in spring when the leaves are out, or that a stolen vehicle tracker has gaps in geographical coverage big enough to hide a car.

When it comes to home healthcare or lone worker protection devices – reliability becomes an altogether more serious consideration.  Solution providers need to be confident that their devices can connect at all times – without steering or preference, to the network offering the strongest signal, whatever the geographical location.

A device that is supposed to work in over 200 countries has to live up to that promise. Take fleet telematics, for example. Logistics vehicles which traverse continents can enter several countries in the course of a week. Fleet managers may have thousands of vehicles making these journeys at any time. To keep track of operations there has to be a solution in place that guarantees visibility in every part of each country they travel through. That requires the vehicle’s ‘black box’ to be able to connect to the best possible signal wherever they might be.

Visibility

An IoT business needs to see where their SIMs are, if they are active and how much data they are consuming at any point in the month. This can seem a particularly onerous task when dealing with thousands of devices active across different locations and time zones. The inability to instantly spot unresponsive or aberrant SIMs could result in a business unable to monitor critical KPIs of industrial machinery, or lose track of precious cargo or even their own workers.

From a financial perspective, an inability to monitor connectivity during an industrial IoT project can also be disastrous. IoT businesses need to be able to diagnose and interrogate rogue devices, enabling them to act immediately to prevent data overuse charges.

One company came to us because they had lost visibility of just one SIM card which had been stolen. Because it could be used on a UK operators’ public APN, its misuse ran to more than £15,000 in a few weeks, and the client only knew about it when that month’s bill arrived. With this in mind the thought of releasing thousands of SIMs into the market could seem daunting.

What’s needed for safer IoT deployment is an industrial-grade system that lets you see all your SIMs in one place throughout the month, and which clearly identifies high use, with the ability to set warning levels and cut-off limits to minimise your exposure to risk.

Business costs

If we’re going to make IoT a fundamental element of businesses, we also need rethink the existing inflexible contractual model. When we are talking about SIM estates running to six or seven figures, then the standard approach of estimating maximum possible usage and buying a large data bundle to avoid punitive overuse charges has to end.

You wouldn’t pay for what you don’t use in any other area of business, so why is this practice commonplace with IoT?

Further inflexibility is introduced by contract terms of 12, 24 or even 36 months that “click-over” at the end to renew for a further full term, with absurdly expensive release clauses should the SIMs no longer be required at any point in the contract. While this is fine in some cases such as AMR and telemetry, for projects like UBI it can create a large overhead burden.

This behaviour has to stop, otherwise it will impact planning for large scale IoT projects which will be seen as high risk, complex and unaffordable. However, large sections of the market still seem to be indoctrinated with the idea that you can only buy data bundles based on predicted maximum use and on long fixed term contracts. This is possibly because the individuals leading IoT projects are unaware there is another way.

This article first appeared in Informa's Cloud & Enterprise Tech Series - http://blog.cloudworldseries.com/2016/09/20/connectivity-often-the-least-considered-element-of-iot/