TPRC45

The need for connectivity and communication during emergencies has spawned research and development of Hastily Formed Networks (HFNs) (Denning 2006; Tornqvist et al. 2009; Nelson et al. 2011; Lundberg et al 2014). These networks, both organizational and technical, are crucial to response effectiveness but are difficult to manage and deploy.

Research to improve HFNs has focused on organizational and interpersonal aspects, including HFNs’ ‘conversation spaces’ (Denning 2006), as well as the technical aspects of designing and deploying network infrastructure. These infrastructures typically focus on connecting emergency response personnel. However, recent efforts are expanding to include connections for affected communities. This expansion of the user base creates new requirements, ranging from different models of network management to far greater diversity in end-user equipment, knowledge and skills.

This paper examines organizational and contextual factors of network design and deployment for affected communities, focusing on (1) identifying the design requirements, (2) decision making concerning the appropriate balance of requirements, and (3) implementing designs. Our analysis also examines the conversation spaces through which decisions, particularly trade-offs, are made.

Through a case study of networks deployed in response to the Syrian refugee crisis, we examine cooperation between an NGO and two tech industry Corporate Social Responsibility (CSR) teams. Together, these partners deployed networks in refugee camps across several European countries. As of March 2017, the system has been deployed at 75 locations, supporting 600,000 users.

As detailed in our case study, design requirements fell into three categories, including replicability, limited management resources, and managing a diverse user base. Replicability generated requirements related to portability of equipment, while limited management resources required networks be largely self-managing. The diverse user base required bandwidth limits to ensure equity by preventing the emergence of ‘super users.’ This was necessary due to network access being provided free of charge. Further, and more importantly, the need for network security affected the balance for each requirement as the increased threat of cyberattacks related to the Syrian war, created an imperative to protect civilians against electronic exploitation.

Preliminary results suggest three outcomes. First, the conversation spaces for network design are positively influenced by ongoing commitments, trust and specialization. The NGO plays a coordinating role as well as serves as a ‘long term player in the market.’ The tech companies trust the work of the NGO’s staff due to their long term relationship. This, in turn, allowed one of the tech companies to focus on the network’s high level design, reflecting specialization between the team participants. Second, the accumulated expertise and an assessment of the context made from a distance by the tech company designers drove decision making concerning the integration of security features into the network’s design. Third, the design task of these hastily formed networks is a multi-stage and ongoing task, with initial decisions made independently prior to deployment, and subsequent decisions made jointly both in the stage of network deployment but also during ongoing operations.