Author: Sanjay Somanath, Alexander Hollberg, Liane Thuvander

Date: April 2021

Subtitle : A platform for collaborative decision making and stakeholder engagement.

Figure : Contextual data fetched from local PostGIS server (Presented at ACT! Sustainability conference - 2020)

Introduction

Digital tools for performance-assessment are commonly used to shorten the feedback loop in testing designs for buildings and neighbourhoods. However, these tools do not extend to the social dimension in the same way as the economic and environmental dimensions.
This project aims to develop a theoretical understanding of social sustainability at the neighbourhood planning level and propose a model for digital tools development. We then implement this model by building a digital platform that addresses the theoretical and operational challenges in pursuing social sustainability at the neighbourhood planning level. The tool is aimed at architects and urban planners, who would then incorporate stakeholder choices and values into the design process.

Figure : A digital tool framework as proposed in Somanath et al. (2021)

Key Features

Collaborative input

Consideration of stakeholder values is crucial to answering the question “social sustainability for whom?”. We propose an information pipeline that allows stakeholders to assess the social sustainability framework and weight their priorities. The SoSu framework comprises of Social equity themes (amenities, community infrastructure, recreation and open space, connectivity, jobs and housing) and Social Capital themes (interaction, participation, stability of the community, sense of attachment and safety).
The stakeholders access the tool over a web-browser and provide their inputs. Bi-directional communication with the design software is made possible through web sockets.

Indicator Selection

Multiple stakeholders can provide their inputs through the stakeholder portal. After the weighting process, the user can select a preferred urban planning framework such as Social Return on Investment (SROI) Methodology, Hyper Locality or Walkability/bikeability. Based on the urban planning framework selected, a list of available indicators inherits the weights from their respective social themes. A planning scenario can then be run through the analytical engine and communicate performance metrics to the stakeholders. Through this whole process, the user can expose additional variables within the design process to the stakeholders.
Figure : Bi-directional communication of stakeholder input and performance output on the web.

Contextual Data and Multiple Data Sources

Over recent years, open data initiatives have made many contextual information layers available to the public. However, this data is located in different locations and different data formats. In this project, we have developed a data pipeline that can link and combine this data and stream it into the user’s design software. Using a guided decision-making process, the user can incorporate multiple datasets from different sources (e.g. Lantmateriet, Trafikverket and SCB) to develop a rich understanding of the local context.

Tool interface

Sanjay Somanath, Alexander Hollberg & Liane Thuvander (2021)
Towards digitalisation of socially sustainable neighbourhood design,
Local Environment,
DOI :10.1080/13549839.2021.1923002
Figure : (Mock UI) Steps for a collaborative digital tool for socially sustainable neighbourhood planning.

Analytical Engine - TripSender

To narrow the project’s scope and refine elements of collaborative-input, ranking, and the analytical engine, feedback from active problem owners is crucial. Problem owners may be any municipalities, developers, building portfolio owners in active development projects, existing or new in the region of Gothenburg.
Our aim with this project is to ensure that decision-making to support social sustainability is not out of a user’s reach due to data availability or technical capabilities.

The generateTrips function from the tripsender module takes data from two a number of live sources and generates trips for all agents in the model.

TripSender Documentation

Read the documentation here

Input

areaName - The name of the area with first alphabet capitalised. type(str)
popPerc - A decimal percentage for the population to be used. type(float)

Output

A keplerGL map type(keplergl.GL)

Output

Run the function and follow the instructions on the prompt. Typing H in the menu will open up a list of neighbourhood names. The names have not all been tested yet though. Note: First alphabet must be capitalised and swedish characters must be used where required. Figure : TripSender Results

Export

The function will also automatically export an index.html once done. This is a keplerGL formatted html with the data hardcoded within the html file. Figure : Figure Caption

Aggregating Data

The results of the tripsender function can then be aggregated at different levels. This is still a work-in-progress.
Figure : (Mock Data) Aggregated amenity demand over a 24-hour period. Figure : (Mock Data) Amenity demand for each amenity over a 24-hour period. Figure : (Mock Data) TCR for male and female agents in the selected neighbourhood

Repo Details

Building the PostGIS server

Coming soon…

Building the project

git clone <repo>
cd <repo>
pip install venv
python3.8 -m venv my_venv
my_venv\Scripts\activate.bat
pip install -r requirements.txt
Open demo notebooks and run all.

Refresh ignored files

Ensure you commit everything before running this.
git rm -rf --cached .
git add .
git commit -m "refreshed ignored files."