Attending a Cambridge Wireless event recently, it’s clear that narrow band IOT (NBIOT) is starting to deliver the coverage and reliability that industrial applications (IIOT) will need. Leaving aside the commercial paths (for instance 5G versus Sigfox) this may follow, by definition a narrow band service will be expensive per-byte.
We’ve blogged and written before about how to squeeze more information (as opposed to data) down this narrow pipe – a range of options centred on information theory or adaptive filtering have been mooted. These were targeted at retrofitting – which is the common situation where a plant operator wants to sweat assets harder by focussing operations on the places that matter and reduce the cost of ownership.
The rapid growth of the IIOT means that the connectivity to operational technology for monitoring and control will soon be designed in. Rather than providing a retrofittable system which needs to learn ‘what is normal?’, the control system can be highly specific to that plant and refer properly to engineering limits.
The Digital Twin
This idea has gained currency – it is commonly called the ‘digital twin’. But is this really new?
Well – no. The problem of ‘estimating’ something which can’t be measured directly has always confronted control engineers — particularly aerospace engineers. Although it’s possible to measure a spacecraft’s attitude more or less directly (with star sightings), velocity is much more difficult. Fortunately, the Kalman filter was discovered in the 1960’s which could be used to estimate the ‘state’ of the craft from the observations which could be made. Crucially, it relied on a digital representation of the behaviour of the craft which was implemented in software.
Control theory and the mathematical methods behind it had become almost a Cinderella profession – ubiquitous in vehicle ECU’s, automated financial trading and biological systems. This is because the tools have advanced so rapidly – what would take many months of analysis can be modelled, simulated and tested with automated tools.
Today’s IIOT is about to become a hugely impressive control system – in which everything from power, transport and water will become part of the ‘state’ and require the right model to predict and decide what ‘normal’ means. The job for control engineers is now to choose appropriate technology and techniques for all this – and deliver business outcomes. The digital twin has truly come of age.