Carl Collins
As we look forward from the 125-year history of CIBSE, it is worth noting the trends emerging now to help us and our members prepare. In this article, I will be looking at how digital engineering is progressing, both in human and computing terms.
Progress on these fronts is becoming more organic, rather than the explosions of change we have seen in the past few decades.
This is to be expected as digital process and ways of working become normalised, and the improvements build upon the step-change advances that have gone before.
What is a digital engineer?
This term has been tossed around for about a decade now, but it is becoming better defined and is merging with the base concept of ‘engineering’.
When was the last time you saw an engineer who did not have a computer? Almost all that we do has some reflection in the digital world, so what makes a digital engineer different from an engineer?
The main aspect is that the digital side is more advanced. For example, a digital engineer will prove a concept by modelling it virtually; a traditional engineer will prove the same by thought, experience and calculation. The two are mutually supportive; we can add the thought and experience to the virtual modelling and create better outcomes.
This means there really is no difference between a digital engineer and the more traditional notion of an engineer. This mends a rift that happened maybe 30 years ago, when computer-aided design (CAD) started to emerge and was just too difficult for most engineers to incorporate.
Before CAD, most young engineers would be ‘on the boards’ to earn their wings before progressing into the ranks of engineers. CAD technicians never had that route available to them, but this is now changing, as the modelling function – which still creates the drawings – also creates the engineering and calculation workflows.
The other side of the ‘digital coin’ is the suite of software platforms that we use to perform the digital engineering function.
Starting in the 1950s, computer programmes have set out to replicate and speed up processes. They have now become very powerful, all-encompassing platforms – so where will the future lie for the essential tools of our trades?
First of all, there has been an inherent mistrust of the calculations done inside these ‘black boxes’. Many organisations have done stress-test trials of software to make sure that they are happy with the results.
CIBSE is now doing this as a service to software vendors with its Software Verification Assessment. This saves hours of testing by customers, and provides a mark of quality that the vendors can use.
Further to this manual checking of calculation outcomes, there is the prospect of artificial intelligence (AI) and machine learning to validate the assumptions and predictions that our calculations generate.
AI can look forward in time by running scenarios, understanding what they mean and recalculating, using revised precepts, to create a holistic solution that has ‘thought’ about more variables than humans can. This leaves us to do what we do best – which is to think of new ways to solve the problems with which we are faced.
The single source of truth
We often hear about the ‘single source of truth’, but how single is that source? Generally, it is at the project level; data exists about a particular project in one place and all the outcomes that require that truth are fed it from this centralised resource.
Yet this is not the whole story. There are some truths that are universal, and this is rarely treated as such. By way of a simple example, the value of Pi can be expressed as 3.14, 22/7 or 3.141592654. These are all correct to various degrees of accuracy, but will produce slightly different results. If the constants were all fed from one source and called up as required, that would remove the variations.
CIBSE is working on a solution for this, by making our data available from a single source that engineers can access by calling them up programmatically. Watch this space for progress reports in the near future!