Connected devices present new opportunities to advance design through data collection in the wild, similar to the way digital services evolve through analytics. However, it is still unclear how live data transmitted by connected devices informs the design of these products, going beyond performance optimisation to support creative practices. Design can be enriched by data captured by connected devices, from usage logs to environmental sensors, and data about the devices and people around them. Through a series of workshops, this paper contributes industry and academia perspectives on the future of data-driven product design. We highlight HCI challenges, issues and implications, including sensemaking and the generation of design insight. We further challenge current notions of data-driven design and envision ways in which future HCI research can develop ways to work with data in the design process in a connected, rich, human manner.

@en

In this work, we develop a vision for entangled ethnography, where constellations of people, artefacts, algorithms and data come together to collectively make sense of the relations between people and objects. This is grounded in New Materialism’s picture of a world understood through entanglement, through resonant constellations, through a multiplicity of unique individual viewpoints and their relationships. These perspectives are especially relevant for design ethnography, in particular for research around smart connected products, which collect data about their environment, the networks they are a part of, and the ways they are used. However, we are concerned about the current trend of many connected systems towards surveillance capitalism, as data is colonised, machinations are hidden, and a narrow definition of value is extracted. There is a key tension that while design, particularly of networked objects, attempts to go beyond human centeredness, the infrastructures that support it are moving towards a less than human perspective in their race to accumulate and dispossess. Our work tries to imagine the situations where participants in networked systems are richly engaged, rather than exploited. We hope for a future where human agency is central to a respectful and acceptable collaborative development of understanding.

@en

A key challenge in carrying out product design research is obtaining rich contextual information about use in the wild. We present a method that algorithmically mediates between participants, researchers, and objects in order to enable real-time collaborative sensemaking. It facilitates contextual inquiry, revealing behaviours and motivations that frame product use in the wild. In particular, we are interested in developing a practice of use driven design, where products become research tools that generate design insights grounded in user experiences. The value of this method was explored through the deployment of a collection of Bluetooth speakers that capture and stream live data to remote but co-present researchers about their movementand operation. Researchers monitored a visualisation of the real-time data to build up a picture of how the speakers were being used, responding to moments of activity within the data, initiating text conversations and prompting participants to capture photos and video. Based on the findings of this explorative study, we discuss the value of this method, how it compares to contemporary research practices, and the potential of machine learning to scale it up for use within industrial contexts. As greater agency is given to both objects and algorithms, we explore ways to empower ethnographers and participants to actively collaborate within remote real-time research@en