Matching the Internet of Things to the pace of the business

I must be a fan of smart connected things – sitting here with 2 wrist wearable devices in a house equipped with thirteen wireless thermostats and an environmental (temp, humidity, co2) monitoring system. However, even with all this data collection, an Internet of Things (IoT) poster-child application that works out the lifestyles of those in the household and adapts the heating to suit would be a total WOMBAT (waste of money, brains and time).

Why? Systems engineeringfrequency response and the feedback loop.

The house’s heating systemhas much more lag time than the connected IT/IoT technology would expect. Thermal mass, trickle under floor heating and ventilation heat recovery systems mean a steady state heating system, not one optimised by high frequency energy trading algorithms. The monitoring is there for infrequent anomaly detection (and re-assurance), not minute by minute variation and endless adjustments.

The same concepts can be applied to business systems. Some are indeed high frequency, with tight feedback loops that can, with little or no damping or shock absorption, be both very flexible and highly volatile. For example, the Typhoon Eurofighter aircraft with its inherent instability can only be supported by masses of data being collected, analysed and fed back in real-time to make pin-point corrections to keep control. Another example is the vast connected banking and financial sector, where there is feedback, but with no over-arching central control the systems occasionally either do not respond quickly enough or go into a kind of destructive volatile resonance.

Most business systems are not this highly strung. However, there is still a frequency response, or measure of the outputs in response to inputs that characterise the dynamics of the system, i.e. the business processes. Getting to grips with this is key to understanding the impact of change or what happens when things go wrong. This means processes need to be well understood – measured and benchmarked.

In the old days, we might have called these “time and motion” studies; progress chasers with stopwatches and clipboards measuring the minutiae of activities of those working on a given task. A problem was that workers (often rightly) thought they were being individually slighted for any out of the ordinary changes or inefficiency in the process, when in reality other (unmeasured) things were often at fault. This approach did not necessarily measure the things that mattered, only things that were easy to measure – a constant failing of many benchmarking systems, even today.

Fast-forward to the 1990s and a similar approach tried to implement improvements through major upheavals under a pragmatic guise – business process re-engineering (BPR). A good idea in principal, especially to bring a closer relationship between resources such as IT and business process, but unfortunately many organisations ditched the engineering principals and took a more simplistic route by using BPR as a pretext to reduce staff numbers. BPR became synonymous with downsizing.

Through the IoT there is now an opportunity to pick up on some of the important BPR principles, especially those with respect to measurement, having suitable resources to support the process and monitoring for on-going continuous improvement (or unanticipated failures). With a more holistic approach to monitoring, organisations can properly understand the behaviour and frequency response of a system or process by capturing a large and varied number of measurements in real time, and then be able to analyse all the data and take steps to make improvements.

Which brings us to the feedback loop. The mistake that technologists often make is that since automating part of a process appears to make things a little more efficient, then fully automating it must make it completely efficient.

While automating and streamlining can help improve efficiency, they can also introduce risks if the automation is out of step with the behaviour of the system and its frequency response. This leads to wasting money on systems that do not have the ability to respond quickly or alternatively, destructive (resonant) behaviour in those that respond too fast.

It might seem cool and sexy to go after a futuristic strategy of fully automated systems, but the IoT has many practical tactical benefits by holding a digital mirror up to the real world and a good first step that many organisations would benefits from would be to use it for benchmarking, analysis and incremental improvements.