Research Fellow I

Singapore University of Technology and Design’s Meta Design Lab (MDL) working closely with the Lee Kuan Yew Centre for Innovative Cities (LKYCIC) is calling for a Postdoctoral Fellow for a new urban analytics project, which will be develop novel urban activity monitoring and analytics with a consortium of influential integrated governmental decision makers in Singapore lead by URA and HDB.

Project Description

This project aims to investigate the relationship between macro-urban configurations and its effect on the activity patterns, social engagement, and wellness of residents, focusing specifically on residential spaces. It hypothesises that specific formulations of integrated functional social activity orientated spaces encourage greater resident engagement, and that these can be effectively replaced elsewhere. However despite demand from researchers, designers, and decision makers, the complex relationship between spatial configuration and activity has not been deeply observed or empirically correlated, not by rigorous empirical observation and data at scale. This research aims to build on existing capability to expand and existing analytical and spatially focused approach, demonstrating and ideally correlating how an urban area design and configuration effects users’ activity and decisions. Expanding novel research into the domain of urban population health research, and related methodological approaches.

To do this it will explore and integrate three concurrent areas of study:

1. An AI driven video image-based object tracking pipeline to capture activity on site data and apply analytics and data visualisation to understand and interpret this.

2. An in-depth ethnographic study of residential location activity and health.

3. Research to understand residential space design its influence and how to improve it to get better spaces specifically in a Singapore context.

The position advertised here focuses on supporting the first section of this and is intended to expand and further develop a data-driven approach to urban analysis, by providing insight into how people use space by using ethical and anonymised image tracking of users of public spaces scaling video and image based ML models and workflows to improve aggregate user monitoring and identification of activity and location. This will be correlated with spatial configuration of the spaces, origin-destination modelling, and available route choice analysis (spatial network) provided though other tracking methods. Then using this empirical data as the input source to do analytics for human insight and potentially exploring machine learning to predict space use, resultant health metrics, leading to automated appraisal of the activity, engagement, and wellness contribution of the space.

This work if proven successful will be scaled to wider analysis of areas where possible to develop data for use in deriving more insight into user activity in tandem with the ethnographic study as data triangulation. Combing both these sources to allow for better informed developing of policy and design guidelines to improve future park and recreation space design, potentially building more user focused ML tools for urban planners and many other agencies and building mangers in Singapore and beyond. The research is highly disciplinary but anchored in Big-Data, Design, Urban Planning, and Applied A.I. the overall project is lead by, Belinda Yuen (LKYCIC), but this part 1 will be headed by Sam Joyce (SUTD), with collaborators from Ministry of Health Transformation, Changi General Hospital, Sing Health, each covering aspects of planning for wellness, applied computational design, A.I., and clinical analysis of spaces respectively.

Researcher Description

We are keen to take on someone who is enthusiastic about conceiving, implementing and testing technologies such as AI, ML, and computer vision to spatial problems, ideally in conjugation with urban analytics and design. The project is intended an investigation into the capability of ML to scale to measuring wellness and design effectiveness in the real world. It would suit someone who is delivery and performance oriented, whilst also being more exploratory and curious about applying novel approaches. The job aims to give a motivated self-starter the research freedom and space to explore new concepts, supported and guided within a group of like-minded researchers. The SUTD environment is that of a start-up university and so well suited to those with initiative both professionally and socially. Someone who takes pride in building quality code and systems to bring new data into understanding the word, and sharing this though papers. The hire will be directly report to Sam Joyce and act as part of the Meta Design Lab how have experience in spatially applied analytics, AI, ML, and image recognition aspects. They will also work closely with LKYCIC in terms of understanding urban activity, wellness, and health aspects and data needs for the user capture and monitoring. This project has support to capture data from government agencies and for potential wider implementation based on findings.

Domain Background (at least one of):

Computer Science, ML/AI, Data-Scientist, Computer Graphics, Computer Vision, Architecture or Urban Planning

Education:

Relevant doctorate (or if expectational a master’s) and significant research experience in one of the above fields.

Desired Skills:

• Experience in applied ML for image recognition and analytics.
• Experience in scaling ML and analytics on real problems with big data volumes.
• Interest in sensor technology, specifically CCTV, remote sensing and related hardware.
• A project lead and outcome driven programming approach.
• Ability to work individually, as well as work with and lead day-to-day small teams.
• A strong ability to work in an open interdisciplinary way with others.
• Interest in architecture, urban planning and design in general.
• Happy to explain and involve non-domain experts in ML and AI.

Required Proficiency:

• Python
• Data science and Analytics
• Experience processing and cleaning data
• Capability in coding TensorFlow, PyTorch or similar ML frameworks
• Experience with server/cloud-based server ML and data processing systems

Desired Proficiency/Interests:

• Analytics and processing of images and/or video
• Big-data processing
• Image Recognition
• Observable Data Visualisation
• Web development frontend and/or backend.
• D3.js and/or Node.js experience
• AR/3D modelling - WebGL/three.js, Rhino/Grasshopper/Dynamo, Unity/Unreal Engine

For any research or position enquiries and submission of application, please direct to: sam_joyce@sutd.edu.sg