I’m Eymen, a PhD researcher from Istanbul Technical University’s Department of Architecture. I’ve developed an app that focuses on optimizing building and housing unit layouts based on the unique needs of different users. This process occurs during the early stages of feasibility and architectural design.
The app link: multi-objective-housing-optimization - VIKTOR
The app generates optimized architectural designs for various housing types. It takes into account factors like views, natural light, sun-hour, sales area, and cost. The goal is to create designs that align with zoning regulations and land specifics.
By trying out the app and sharing your thoughts, you can help refine and enhance it. I kindly ask for a small favor – try out the app and complete a brief survey afterward. Your honest feedback will greatly contribute to my research and the app’s development.
Hey Eymen, this app looks great! A very nice addition to our app gallery . Could you provide a short description, and a great image that we can use for the app gallery?
Hi Eyman!
Super nice app, I enjoyed exploring it. Best of luck for your PhD as well. I have few questions in mind & I made this video so the discussion is visualized.
I find it powerful to have all early design optimizations in one place! In my 2nd attempt I maximized saleable area (profit) and I expected using the whole site plot (red rectangle) for the tower. However, I find other results. Is there another criteria leading to this output?
As an ex-architect, I am curious to know how the solar hour analysis and view are performed.
Hi Julia,
Here is the description:
This app finds the architectural form and layout of the building and each unit within it, according to the needs of user types, at the early stage of architectural design.
I am uploading a picture. I hope it’s “great”
Hi Mostafa,
I had initially used an outdated dataset, which caused the model to perform poorly. I have now updated the dataset, which should lead to better results. Please give it a try. It’s worth noting that while adjusting the parameters might have some impact, the changes might not be drastically noticeable. Here’s my explanation to that and other questions of you.
The model is generating outcomes based on the plot area and FAR. For the plot in the model, we have 1823m² area and 2.07 FAR. In Turkey we also have %30 percent exclusions for that area. So, our sale area should not exceed 1823x2.07x1.30 =4905 m². Model already optimized all criteria given. From 20.000 alternative, users can only see approx. 4000 pareto optimal solutions. When you maximize sale area there is a limit that it should not exceed 4905 m².
I added description for all parameters (view, daylight, sun-hour, sale area and cost). I think it’s a little more descriptive now. Writing of the exact calculation parameters and limitations is in progress…
I should also note that when you select values for unit mix or user preferences, I sum up all the values for that category. Then, I scale and proportion the values you’ve chosen for the parameters. This is because I am not sure how to restrict the user and modify or limit the other parameters accordingly on Viktor platform.
For example, giving 10 points to all units and giving 1 point to all would yield the same result. The best practice is to allocate a total of 10 points for unit mix and another 10 points for user preferences. Adding another note about this scenario: due to the large size of the land, even if you choose higher proportion of 1+1 and 2+1, the model will include as many 1+1 and 2+1 units as possible however, it might not be feasible to fulfill all these preferences exactly as you want.)
. Could you provide a short description, and a great image that we can use for the 
