As computer systems become more sophisticated and capable of performing complex technical tasks, what impact will this have on professional engineering? Specifically, engineering calculations.
As computer systems become more sophisticated and capable of performing complex technical tasks, professional engineering may undergo changes – similar to those that greatly reduced the roles of skilled craftsmen during the industrial revolution and manufacturing trades during the 20th century.
During the industrial revolution, highly skilled craftsmen were largely replaced by mechanised processes overseen by tradesmen. These mechanised processes have now been computerised, replacing those tradesmen with fewer technicians.
This article is the second part of the two part series considering future trends in engineering. Part 1 considered trends in draughting
and identified barriers preventing adoption of integrated 3D CAD in some sectors. In Part 2, I consider trends in engineering calculation and convergence of draughting and engineering into common automated systems.
Historically, engineering design has been an iterative process. Normally this involves three or four iterations of each engineering specialisation revisiting the design in greater detail, culminating in preparing or updating a capital cost estimate. The typical stages for an industrial project are shown below in Figure 1.
Figure 1: Typical industrial front end loading (FEL) iterative project development methodology (the particular disciplines and sequence will vary from project to project according to the project scope and design dependencies)[/caption]
At the completion of a design-iteration, the design and estimated cost are compared against the project objectives/business case. Minor modifications in scope are made and the project is subject to further expenditure being approved before the next design iteration is started. The early iterations are at a high level only, with each successive iteration being more detailed, time consuming and expensive. This approach has been very successful allocating expensive and time consuming engineering design to the development of acceptable solutions.
Moving through disciplines sequentially in each iteration allows the work to be procured as separate, single discipline engineering services engagements. In some industries there has been a benefit to having some overlap between disciplines, either through a multidisciplinary engagement or collaboration between single discipline providers. However, clients balance this against the reduction in competitive tension from reducing the number of potential service providers to only larger multidisciplinary consultancies.
Computer Assisted Engineering
Engineering computational tools are widely used within different engineering specialisations. They range from simple spreadsheet calculation templates, to complex finite element analysis systems for modelling stresses or fluid flows. These are becoming more sophisticated within increasing capacity of affordable computers, software development, and the continuous improvement of engineering and design standards.
There are currently a number of initiatives underway to integrate these specialised offerings into more comprehensive multidisciplinary modelling systems. An area of particular activity is to add these engineering tools into common 3D CAD platforms. This takes advantage of the software development capability within CAD software vendors, and existing relationships and familiarity between these vendors and many of the different engineering specialisations. CAD software vendors are both developing these new engineering systems in-house, and through the acquisition of specialist engineering software developers.
As computer assisted engineering design software becomes more capable, it will be reduce the number of engineers required to complete engineering calculations in the same way that CAD reduced the need for draughting ‘tracers’. There will still however, be a need for experienced and technically capable engineers to select the appropriate software, configure it to the particular application, select inputs, verify outputs and interpret these outputs for others.
Using computers in place of manual engineering calculations reduces the engineering man-hours required per design iteration. Reducing the engineering man-hours/cost per design iteration is likely to increase the number of iterations used and the benefits of overlapping or concurrent work within different disciplines. This is likely to accelerate the consolidation of single discipline consultancies into multi-discipline consultancies.
It will also reduce the number of back office engineers required to complete engineering calculations. Although the numbers are unlikely to be large, there will also be a shift of some highly trained and experienced engineers from consultancies to the development of engineering software. This will reduce the overall cost of engineering design, and shift some of the remaining cost from engineering consultancies to engineering software vendors.
While continuing to favour multidisciplinary consultancies, fewer engineers will be required. There will be greater reduction in manual engineering calculation by junior and intermediate engineers, with a lesser reduction in the number of experienced and technically competent senior engineers. These senior engineers are currently trained though junior and intermediate roles, exacerbating existing succession issues within the engineering profession. This may be somewhat addressed through graduates undertaking additional postgraduate study prior to joining the workforce.
The development of computer assisted design software will be quickest in markets where projects are numerous, standardised and relatively simple. Design of numerous systems offers a greater number of opportunities for the deployment of software, while standardised and simple systems are easier to automate.
Engineering Design Automation
Over time, these integrated design systems will become even more sophisticated and automated. They will be operated by smaller teams and eventually without human technical oversight. The combination of automated design systems that have a greater scope of capability, while simultaneously delivering a more reliable output, will drastically reduce the number of engineering man-hours required to prepare a design and thus the cost per design.
While it may appear that obtaining technical design inputs will still require professional engineers, even for a fully integrated and automatic design system, this requires further examination. Design automation is likely to be mirrored in other fields such as gathering geotechnical information. Eventually projects may simply reference a comprehensive database of automatically gathered geotechnical information for a region rather than requiring the services of a geotechnical specialist. Similarly, district planning rules are likely to be codified and accessed automatically by the design software.
The combination of automated design systems that have a greater scope of capability, while simultaneously delivering a more reliable output, will drastically reduce the number of engineers required to complete a design. Identifying user requirements and client interaction are likely to be the most difficult areas to automate, and engineers in these areas are likely to have greater job security. This will be particularly relevant for projects with multiple stakeholders or where the project has aesthetic or artistic elements that are difficult to codify and automate.
Design for niche projects is likely to be more slowly automated. And design for novel projects is unlikely to be automated unless they become routine.
Typical engineering consultancies, or whatever the entity fulfilling this role is called, will be small and multidisciplinary organisations that are primarily client facing. A small back office component will be responsible for selecting software, managing the relationships with software vendors and managing the flow of information into and out of design systems. These may run on hardware owned by the consultancy, or more likely on that owned by the software developer providing engineering software solutions.
Engineering consultancies specialising in novel or niche projects are likely to be changed more slowly by design automation. They may make judicious use of automated tools for some parts of a design, but will remain mindful that such tools will not be able to complete all aspects of the design, and will need expert technical oversight to confirm that validity of the parts they are used for.
The trends outlined are summarised below:
||Automated and integrated engineering systems
||Integrated multidisciplinary systems
||Engineering and draughting teams with a range of skill levels
||Engineering and CAD with a reduction in ‘tracing’ and junior engineering positions
||Engineering and shared CAD with less back-office engineering personnel
||Client facing with small engineering and IT systems support team
||Small single disciplinary consultants and large multidisciplinary government engineering departments
||Large multidisciplinary and specialist single-disciplinary consultancies
||Smaller multi-disciplinary consultancies
||Small ‘shop front’ multi-disciplinary consultancies
This is both an exciting and frightening time to be an engineer. On one hand, our traditional positions will be increasingly automated and able to be performed by fewer people. Many of us that are currently in these roles will need to find new jobs, which may require retraining, taking a more junior position in another existing profession, or moving into an emerging technology.
On the other hand, we are on the cusp of being freed from much of the engineering drudgery. This means we will be able to give more attention at the start of the project to fully understand our clients’ requirements, consider innovative solutions and use a greater number of design iterations to develop more optimised solutions.
Our industry and engineering educators will need to improve the training of new engineers, because in the future, they must achieve a high level of technical proficiency quickly rather than incrementally during their progression from junior to senior positions over a number of years.
In part one of the series I consider the trends in draughting and identify barriers preventing adoptions of integrated 3D CAD in some sectors. Read part one here.