Social Drones

Autonomous drones are on the cusp of becoming ubiquitous in and around our built environments. They are already used widely in contexts including construction, planning, energy, geology, archaeology, military and law enforcement, search and rescue, logistics, forestry, photography and filmmaking, and real estate.

Most current drone applications involve a pilot using a drone to remotely view or manipulate things in destinations that are not easily accessible for humans. Conversely, there are emerging applications where fully autonomous drones operate in spaces populated by human users or bystanders. We coined the term "social drones" to describe these applications.

Social drones and “human-drone interaction” (HDI) are growing topics within human-computer interaction (HCI) and human-robot interaction (HRI) research communities. Situated here, we conduct design research on interactive autonomous drones.

Our work develops human-centered knowledge and design experiments around social drones. The foundation of our methodology is "research through design", where we undertake design experiments as follows:

1. We build prototypes for social drones and related systems (e.g. information and interaction), informed by previous research. Our designs are sometimes pragmatic and task-oriented; but more often, they are speculative and playful.
2. We conduct human-centered empirical design studies to characterize experiences around these inventions.
3. We analyze and theorize over the results to extract coherent frameworks that can inform future products, services, policy, and research.
4. We publish our findings, as well as open source software, tools, documentation, and other contributions to support future work.

This work is broadly consequential, for four reasons:

1. The usefulness of current drones in specific work and leisure contexts is indisputable, yet they come with significant usability and user experience limitations for both users and bystanders. Research and experiments to conquer these will serve existing stakeholders and enable novel use cases.
2. Drones are already popular, and already capable of harming and disrupting our lives if things go wrong. Social constructs (etiquette, law, policy, education...) are already evolving to cover them, and we would like to support this evolution with research and knowledge.
3. Knowledge on interactions with autonomous drones can inform designs for piloted drones, which are widely used for commercial applications. For example, assistive features like autonomous obstacle avoidance must be designed by foregrounding human factors related to both pilots and bystanders.
4. We expect social drones to become ubiquitous in our built environments, in which case the scale of consequences justifies investing in research.

Many machines and technologies -- operating with various levels of autonomy and “intelligence” -- are already embedded in the fabric of our societies, and a variety of social patterns (norms, conventions, expectations, etiquette...) accompany them. The busses and automobiles on our streets, the computers on our desks at work, and the tools in our kitchens mediate and influence interpersonal relations and social phenomena. We expect drones to become similarly ubiquitous. A sophisticated understanding of how drones interact with human factors and social order will be key to useful product designs, and to the deft handling of drone-related phenomena in the public sphere.

Selected Results