Tomorrow’s Architect: Unveiling a Data-Driven Vision for 2034.

It’s 2034, 10 years from now.

I’m sitting at my desk at work and I get a call from a client I’ve worked with before. My client tells me that they want a new, larger data centre to be designed and built, however, the design timescales are short. In addition, they will compare the timescales and prices from two other architects who also specialise in the design of data centres.

They also want the full (RIBA) service, stages 0 to 7 and as such need a fee quotation for all stages. Two years ago we delivered a Data Centre for this client, in addition, we also have experience with a couple of other data centres for different clients. An hour later I receive an email from the client containing the brief, EIR’s, and their updated standards/requirements. They have also attached a digital topographical site survey of where it is to be built.

This is where the magic happens.

I drop the brief into my AI Document Intelligence service which applies advanced machine learning to extract text, key-value pairs, tables, and structures from documents automatically and accurately. This AI service turns documents into usable data and shifts your focus to acting on information rather than compiling it.

I ask the AI system to highlight the differences between the last 3 briefs and to give me a report. In seconds I understood that this Data Centre is twice the size of the last one but also requires a higher level of physical security than previous projects, an important requirement not to miss. The other challenge is that energy reduction has to be reduced by 10% as the company’s targets on sustainability have become more challenging.

I ask the AI system to analyse the design constraints and parameters, including site geo data and also prompt the AI to analyse the following:

1. To take the 3 previous projects “As Built Models’ that were all designed in BIM to level 3 and to create new options.

2. I ask the AI system to take the cost data from the last 3 projects and link this to live cloud pricing data to give me a cost for each option.

3. I ask the AI to search our company databases on the fees that were agreed and give a price taking into account today’s inflation. I ask the AI system to build in a 20% profit.

In less than twenty minutes my AI system presents me with 5 design options, fully costed and gives a breakdown of fees for all RIBA 7 stages.

I send my fee proposal over to the client and wait. I do not send my options for the 5 designs.

A week later (clients are rather slow) I receive a call from the client informing me that we have won the project and they apologise, stating that the other architects had been slow at calculating their fees. The client tells me that they want to proceed but want some design options for a meeting in a couple of weeks.

Now I am ahead of the curve, I could send over the fully costed options that our AI system has generated but I take this time to make some informed tactical decisions.

I look at the designs through the eyes of a human architect and go about optimising the design of the building, considering the pro’s and con’s of all 5 options and their positioning on the site.

Once this is complete I present 3 options to the Client. The client chooses his favourite, guided by myself, the architect. The client then insists that the building must be delivered for the budget, this is critical. In the back of my mind even though we are in the concept stage I am confident that the AI data supports any option they could have chosen.

We get the client’s approval for the project design and prepare our planning submission.

We then use AI to run a code/regulation compliance verification for the planning application to demonstrate that the proposals comply with all local and national standards. We also run an AI program for design risks which informs our Designer’s Risk Assessment (DRA) and identifies any residual risks.

We plug in our environmental software, including data from the last 3 projects to demonstrate that we meet the energy requirements. We calculate our embodied carbon and run our BREEAM and LEED compliance checkers and the design is given a target rating.

Since 2031 AI systems can understand all construction materials on the market and this now gives us powerful data to down-select materials based on performance criteria such as; thermal, fire and design life (too many criteria to mention here).

We submit for planning in the full knowledge that our design is compliant with planning requirements, regulations, codes, country standards, environment compliance and energy consumption based on our AI systems.

Planning is achieved in the 8–13 week time frame, during this period we have put our Architects to work on 16 other projects of a similar size, we want to make money after all.

With Planning consent now gained we are instructed to proceed to prepare construction drawings and specifications.

Our AI system scans our database, offering up details from past projects but has the parametric ability to adjust for different materials and offer prompts and warnings when the architect needs to make a decision based on design aesthetics. The AI will always offer a solution that complies in terms of staying inside its technical parameters and relationships with other material are also factored in.

The specification is linked to the drawings and instantly changes if materials are swapped out or if the BIM/digital model material is deleted or revised. Live cost variations are used which inform decisions on the selection of building elements.

Architects are now making data-driven decisions.

The construction sequencing is worked through with the contractor who has been engaged within the framework agreement but this could be done after a successful tender, the earlier engagement with the contract the better. We run constructability sequences, asking AI to suggest crane positions, crane movements and steel erection sequences.

We have also identified materials that require long lead times and our contractor is beginning to place orders. The contractor is impressed with our insight and understanding of construction and we begin to collaborate on a deep level.

The project moves to the construction phase at site. The contractor links to our model data and sequences the whole of the build, live weather data informs adjustments to daily planning and requests for information are TQ’s are managed centrally through an online platform linked to the digital contract. Emails become workflows and all parties can be involved in key decision-making, including the client. A culture of openness and inclusion creates project satisfaction for all.

Drones are used in conjunction with agile robot dogs are used to augment the workforce The robots are fitted with LiDAR which patrol the onsite build and give the architect real-time data (RTD) on progress. Autonomous drones highlight safety issues, especially at the site and send the construction manager alerts.

The building is now complete, delivered ahead of time and under budget and profit targets are achieved. The As-built model is completed and handed over the the client’s asset management company to help operate the facility, a digital AI twin.

A post-occupancy evaluation is carried out and essential client feedback is taken and assessed and fed into the AI system. AI will now help deliver a culture of constant improvement and learning for future projects.

Isn’t 2034 going to be awesome with AI as your co-pilot?

Authors comments: Improvements — A client could let us live link to the asset data so that we could have access to the historical data to better understand performance issues, maintenance issues and operational issues. This means we could run the AI and optimise the design further.