Relief Applications

ILO - Working Conditions Data Collection Chatbot

The Vision Zero Fund, working with LABADMIN/OSH, wanted to identify occupational safety and health deficits across Mexico's coffee value chain, reaching workers standard surveys might miss. We built a chatbot on Facebook Messenger with a visual, low-text interaction that lets workers share data anonymously, plus a website monitoring tool for team members.

Client
ILO
When
Oct 2020–Mar 2021
Team
3+
Services
Info management & MEAL, AI tools
Technologies
Node.js, Microsoft Bot Framework, MongoDB, Angular
Industry
Fundamental Labour Rights
SDGs
8. Decent work and economic growth, 10. Reduced inequalities
Client website
ilo.org

Mexico has experienced challenges in gaining access to timely information about different issues,associated with coffee production, inside its territory. Despite technological advances in information capturing and systematisation Mexico has no updated coffee production registry.

Vision Zero Fund (VZF), in coordination with LABADMIN/OSH, wanted to develop a crowdwork-based public platform to identify and visually represent OSH deficits throughout the national coffee value chain. VZF's goal was to achieve zero severe and fatal work-related accidents, injuries and diseases in global supply chains.

To reach this objective the best strategy was to develop a chatbot. In addition to collecting information, the chatbot could direct respondents towards information resources or external grievance mechanism portals. It would need to enable all workers to anonymously generate and share data, thus spotting problems at scale.

Relief Applications team chose to use Facebook Messenger as a platform to generate data from those, who might not be reached through standard survey methods. This led to a better understanding of the work conditions and helped identifying the main occupational safety- and health-deficits.

The challenge

  • Develop and deploy chatbot on a system widely used by the beneficiaries
  • Make interaction accessible for any user
  • Lack of visual support

What we did

  • Chatbot deployed on Facebook Messenger
  • An appealing chatbot model based on visual interaction and less writing
  • Data visualization monitoring tool accessible on website for team members
Working Conditions Data Collection Chatbot, ILO: screen 1

More projects