These days, the ability to acquire large quantities of data is easy but insightful manipulation of data requires expertise and is a lonely task. We studied experts and novices in the fields of genomics, synthetic biology, neuroscience, and botany to identify key workflows where collaboration might be beneficial.
To understand our users, we studied their existing workflow in situ, interviewed novices and experts, and tried to become our users via immersive activities and workshops. Once we had a better understanding of their process, we collaborated with our users during ideation and brainstorming sessions, identifying potential opportunities for improvement that could be addressed with technology. Following the collaborative brainstorm, we began the divergent design process, creating many prototypes that might address the problem and sending them to our users for feedback. We iterated on the prototypes with our users’ blessing and reaped what positives we could from those that did not. Increasing the fidelity and regularly requesting feedback from our users, we arrived at something we could evaluate in the wild.
Scientists that must go out into the literal field to collect data have been using paper and pencil plus miscellaneous instruments to record their measurements, a separate camera, and voice recordings that would later be combined via the potentially error-prone method of syncing date and time and through spreadsheets. Identifying the needs of these scientist and the general workflow, we created a mobile app that consolidated all the heterogeneous data into one data set automatically at data collection time. This assisted in standardizing the data set and guiding novices with additional information as needed as well as definitions of unknown terms. The next problem was analyzing the data. The existing method as discussed above is to go through the spreadsheets, which is difficult when working in pairs. To remedy this, a tabletop interface was created that provided a richer grasp of the data set for novices, as well as, democratized the data such that the students were able to collaboratively gather insights from the data set.
Once the issue of data-driven collaboration was somewhat address, we went on to tackle the issue of collaborator scale. One of the limitations of existing tabletop interaction is the two person limit on the device due to data orientation and scale. To address the scale issue, we designed and implemented a large scale interactive tabletop for augmenting data-driven collocated meetings of larger teams (8-12 participants). Once built, we encountered numerous usability challenges for large-scale tabletop interaction and devised a set of design metaphors and interaction techniques for supporting collaboration around large-scale interactive tabletops. The results of this work can be found in Enhancing Data-Driven Collaboration with Large-Scale Interactive Tabletops, ACM CHI 2013 Workshop on Blended Interaction, ACM CHI 2013.
Next we asked ourselves, but what about non-collocated collaboration? What about a more tangible interaction? We started generating designs and exploring many options but we found that our novel designs needed validation at a gestural level. We began with paper prototypes and increased the fidelity to full one smart tokens and a tabletop interface. The results of this work can be found in Enhancing Data-Driven Collaboration with Large-Scale Interactive Tabletops, ACM CHI 2013 Workshop on Blended Interaction, ACM CHI 2013.