Skip to content

Introduction

Jump to bottom
zoeps edited this page May 28, 2021 · 16 revisions

Goals

The INTErventions, Research, and Action in Cities Team (INTERACT) is a pan-Canadian collaboration of scientists, urban planners, and citizens uncovering the impact of urban changes on health and equity. Our aims are:

1. Understand Context: We use concept mapping to capture the vision and perceptions of diverse stakeholders, helping to drive successful implementation of change.

2. Measure Change: Citizens share data through online questionnaires and mobile technology, while stakeholders feed our geographic information system with data on urban change, allowing us to track the impact of an intervention over time.

3. Analyze Impact: Using quantitative and qualitative methods, we conduct longitudinal analyses to determine the impact of urban form interventions on physical activity, social participation,well-being, and related health inequalities.

4. Mobilize Knowledge: INTERACT supports stakeholders with evidence and recommends local actions to promote the design of smart, sustainable and healthy cities for all.

Describe overall project. Keep it tight and high level, with links to other sources to provide more detail. Here we should focus on issues that inform the decisions taken within about the data processing itself. ($zoe? $jeff see my additions above)

Roles and Responsibilities

Responsibilities within the project are divided among three basic role types:

  • INTERACT works with several data partners who provide services for collecting metadata, telemetry data, and/or survey data from participants
  • Each city in the study has a research coordinator who spearheads local recruiting of participants, handles device assignments, metadata collection, telemetry data collection, and provides liaison between participants, data partners, and other project team members as required.
  • Within the project team, there are several data wranglers who work with coordinators and data partners to ensure smooth migration of the collected data to ComputeCanada and then oversee the verification, cleaning and integration of the migrated data into a form suitable for researchers.

Data Pipeline Phases

Studies within the INTERACT system follow a data flow that divides the process into a number of sequential steps:

  1. Recruitment: in which participants are invited and enrolled into a study
  2. Collection: in which participants generate digital telemetry of their movements and answer surveys/questionnaires
  3. Migration: in which the collected data is moved from the collection points to ComputeCanada
  4. Ingest: during which the migrated data is verified, cleaned, and normalized into usable raw-data tables
  5. Digest: in which the raw data is summarized and annotated with basic metrics to prepare it for analysis by researchers

Data Pipeline Infrastructure

Describe the resources used to assemble the system: 3rd party partners, institutional services, key technogogies. ($jeff,$zoe)

Data Collection Partners

  • SenseAnalytics provides INTERACT with SenseDoc devices, which are hip-worn tracking devices that cache their telemetry data locally for later retrieval by our study coordinators
  • Ethica uses a smartphone app to collect telemetry and survey data from registered participants and cache it on secure servers until it can be migrated to ComputeCanada upon completion of the study
  • Polygon (formerly known as Treksoft) provides our coordinators with a participant management portal and also manages and hosts a survey platform, including the VERITAS tool, a map-based questionnaire combining social and spatial networks
  • ComputeCanada is a national research network that provides advanced research computing and storage services for Canadian academic researchers

Software Systems

Describe the types of servers, software, programming languages, and data technologies used in the pipeline ($jeff)

  • servers: all linux-based
  • applications: SenseDoc extraction tool, exactfile fingerprinting tool
  • data formats: CSV for Ethica data, protobuf (Ethica data transfers between phone and caching servers)
  • programming languages: primarily Python 3 with some bash scripts
  • database: PostgreSQL, sqlite3, Cassandra, MySQL? (ask Benoit re Treksoft db)
  • continuity tracking: our internally created ProvLog system

Clone this wiki locally