The Challenge

The challenge that Village 2047 presented us was a unique and diverse one. While envisaging for the villages at hundred years of independence it encourages us to think of the future while addressing the problems of today. The task was to come up with a solution for mass housing for a low-income group. The design solution had to be such that it could be replicated in different parts of Indian villages, where most of these villages lacked skilled labor.

The scarcity of skilled labor allowed us to come up with a precast and assembly module setup. With the help of technology, we decided to write an algorithm that could work according to the conditions of the site and lifestyle in a particular part of the country. To develop this example, we took Koth, Gujarat, as our site and researched the patterns in their habitat. Based on these observations we created a logic through evolutionary algorithms that would incorporate factors in design like physical context, cultural variation, etc. that would give us design iterations. We used genetic algorithms as an optimizing algorithm.

Our Perspective

The scarcity of skilled labor allowed us to come up with a precast and assembly module setup. With the help of technology, we decided to write an algorithm that could work according to the conditions of the site and lifestyle in a particular part of the country. To develop this example, we took Koth, Gujarat, as our site and researched about the patterns in their habitat. Based on these observations we created a logic through evolutionary algorithms that would incorporate factors in design like physical context, cultural variation, etc. that would give us design iterations. We used genetic algorithms as an optimizing algorithm.

Design Process

The evolutionary algorithm consists of two parts. One is developing the generating algorithm and the other is evaluating algorithm. The task of generating algorithm is to generate various iterations based on overarching logic and design inputs. The task of evaluating the algorithm is evaluating whatever is being produced by generating algorithms and in the process optimizing it.

Developing Generating Algorithm

Every house in the village has a distinct character to it. On closer observation, one realizes that there is an underlying pattern/structure that each of these houses follows. For decoding this structure, we first removed the variance and then classified them into a kit of parts from which the building is made. This kit of parts consists of staircases, toilets, semi-open spaces, closed spaces, and balconies.Once we had a kit of parts ready, that constitutes original houses, in the second step we wanted to the interrelationship between these. It was not possible to study each and every house. Hence a clustering algorithm was made which could crawl on the site map to find interrelation maps that were found in most of the houses in the village. Once data collection is done, the next step is to make the generating algorithm. The generating algorithm was made in such a way that it took elements from the kit of parts identified earlier and using interrelation maps formed the house.

Evaluating Algorithms

The designs that we generated by generating algorithms were evaluated based on the following parameters.

The amount of solar radiation received in semi-open spaces.
The amount of solar radiation received in closed space.
View towards places marked. (maximized)
Wind flow in closed spaces. (maximized)
The number of people living per unit area. (maximized)
Area of Semi-open spaces to the area of the footprint of the building(maximized).
Sky view factor.

To find out the amount of solar radiation that needs to be attained in semi-open spaces and closed spaces, several clusters were studied from the village and a range was obtained. The same was done for the sky view factor. Based on the parameters we got, an evaluating algorithm was made. The script also made sure that the building design finally achieved could be made from precast modules.

Final Stage

Through the genetic and evolutionary algorithms, we got the corpus of the building. Since one of the challenges in this project was that there are very few architects and engineers that work in these rural conditions it was important that the construction process should be simplified in a way that after the initial phase the community can take over the growth of the project.In order to build the structure, we designed a method through which the construction process would take the least amount of time and most importantly require little or no skilled labor. Hence, we developed a modular system that consisted of three types of structural modules, one module for flooring, and one module for plantation in the semi-open spaces. The core consists of a staircase, elevators, and services that are built (in-situ). The rest of the spaces both in the semi-open areas and houses are built using a modular system. This enables the possibility of growth of this housing by the community using it.Once these houses are built, different users will come and appropriate these houses according to their needs. This will give a unique character to each and every house. After going through the design process we came full circle by comparing it to the existing typologies.