Enabling Automation in The Built Environment

Enabling Automation in the Built Environment

The building and construction industry is primed for automation. The automatic manufacturing of the “stuff” of a building – the parts and pieces that constitute the building itself – is already a matter of course. We hardly blink an eye at automated production lines, pre-fabrication, or even using CNC technology to produce structural members. Moreover, as the fourth industrial revolution continues at a breakneck pace, the ability to automate even the manual tasks that assemble this building “stuff” is becoming less and less of a fantasy. Indeed we are seeing rapid advancement in automated bricklaying, 3D-printed buildings, and excavation. By analyzing the analog task, developing algorithms to optimize the tasks based on a set of parameters, and creating or retrofitting a machine to execute this algorithm, the assembly of building “stuff” can be achieved in record time.

At VIATechnik, the question we often ask ourselves is what if we could take this one step further? What if, in addition to automating building “stuff,” we could automate building “ideas” – in other words, what if we could apply similar data analysis, optimization, and execution methods to automate our thinking about buildings? By using generative design principles and computational BIM, our team is actively working on a variety of projects to automate and optimize the Built Environment.

Generative Design

Generative design mimics the evolutionary processes found in the natural world. Given a set of input parameters and prioritized design goals, a generative design program can rapidly iterate all possible permutations of a design problem. The solutions that best achieve the initial goal set are then re-combined and re-iterated as the program learns what works, while the solutions that do not achieve the goal set are eliminated. Positive attributes are selected until an optimized solution is reached.

Consider the following problem: a client is choosing between several new sites for development. The analog approach to this problem is to collect existing zoning data; determine how much overall development each site can hold; prepare layouts for each site to balance leasable area against parking ratios, cost per square foot, FAR, zoning, and other constraints; use these layouts to determine ideal building use; then return to the market data to check feasibility. Rinse and repeat as necessary. With generative design, all of these parameters and more can become inputs into one model that produces optimized solutions for each site based on the priority assigned to each parameter. For example, it can show you which site is best for a light manufacturing building of a given construction type with a certain number of loading dock spaces. Or say you have a tax break that only kicks in if a certain amount of space is left undeveloped on the site. A generative design model can show you not only which site and building types are optimal for achieving this credit, but how to distribute that space ideally on the site given environmental or zoning constraints.

A system such as this has broad applications beyond the scenario described above. Generative design can be deployed across multifamily residential, healthcare, industrial, mixed use, student housing, and commercial/corporate projects. Its constraints are unlimited: it can take into account constructibility, cost, code regulations, material variance, building use, structural systems, and time just to name a few. It is also scalable, able to tackle problems of any size from building form to panelized cladding to interior fit-out to object design. Achievable through a wide variety of programming languages and plugins, including Grasshopper, Node Box, Revit/Dynamo, xGenerative Design/Catia, and Dreamcatcher, among others, generative design is the next step in evolving how we think about architectural problems.

Computational BIM

Building information modeling (BIM) has become increasingly implemented across the building and construction industry. Automation hinges on the leveraging of information, and since information is right the in the name, BIM is ideally suited to help automate tedious coordination and notational tasks. Using Dynamo, pipe sleeves can be added at floor penetrations, tags added to all ends of pipe, and bottom of pipe data extracted for populating coordination drawings – all automatically. With Synchro, many scheduling possibilities can be visualized and analyzed in a 4D BIM environment. In this way, construction time can be optimized in the same way as the building itself. Still another dimension – cost – can receive the same treatment with 5D BIM technology, turning budget calculations into another optimizable problem. By utilizing the above technologies, the VIATechnik team is able to create a digital thread through all phases of construction – design, manufacturing, engineering, and installation. This will one day enable a fully integrated digital communication network that will improve efficiency, reduce costs, and allow the industry to more readily adapt to changing technology and market forces.

Conclusion

Automation in the building and construction industry is here. As technology, programming, data collection, and artificial intelligence continue to gain widespread implementation, more and more of our ideas about building can be automated. This may seem like it robs us of our essential agency in the process – on the contrary, if we can rearrange the way we think about the idea of a building, we can become the designers of the way we solve the problem and leave the solving (where possible) to the computers.