Jennifer Taylor, Stephanie Sherman
Human-centered design (HCD) is championed as a means to better understand the activities, goals, and interactions of people with technology, toward designing solutions grounded in the right problem or need. We sought to explore how HCD could be integrated into events that focus on tackling complex problems but currently have a limited user perspective. We focused specifically on hackathons, as they provide excellent forums for mobilizing and leveraging technical skills toward a shared purpose, but often lack critical capacities drawn from HCD. Solutions developed at hackathons, while technologically impressive, may lack deep understanding of the problems that users encounter, or may be unfit for the communities they intend to serve. Design-a-hack-a-thons are intensive design events that integrate HCD components into the typical hackathon model, giving teams methods for focusing on the needs of people in real-world challenges while they develop technical solutions.
The Design Lab at UC San Diego and the MIT Media Lab City Science group  co-produced a three-day design-a-hack-a-thon on the theme of people-centric mobility in Cambridge, Massachusetts, building upon the City Science group's experience of running hackathons and the Design Lab's experience in supporting HCD research and education. The event brought together interdisciplinary college students alongside scientists, engineers, designers, and policymakers from the public and private sectors to address emerging urban mobility challenges. The following describes the methods and values of this first design-a-hack-a-thon and offers recommendations for orchestrating such an event.
Design-a-hack-a-thons draw inspiration from and share similarities with several related intensive design-event models, including designathons, design competitions, and design swarms [2,3,4,5]. These models deploy different approaches, time frames, and levels of guidance from coaches or organizers; all, however, illustrate the power of convening a variety of stakeholders to understand and address societal challenges by applying HCD. The design-a-hack-a-thon model focuses on integrating HCD methods like user research, user experience, problem finding, and iteration with the technical expertise mobilized in typical hackathons. Because hackathons provide an existing and popular model of co-production, they present a valuable opportunity for introducing HCD to a wider audience.
For our first design-a-hack-a-thon event, we focused on mobility because it is a complex system that poses social and technical challenges for various user groups. The prompt for the People-Centric Mobility Design-a-Hack-a-Thon asked: How can new methods of mobility—personal or shared, passenger-use or goods-moving—be designed to prioritize the human experience? Three thematic areas of exploration—life mobility, socially intelligent robots, and the new street—guided participants toward challenge areas.
Our design-a-hack-a-thon took two months to organize, promote, and produce; the relatively short preparation period required was due to the MIT Media Lab's prior experience running hackathons. We hosted 250 people, including student teams, mentors, and organizers, for an intensive, three-day event. The first day included orientation, team formation, and problem identification; the second day involved observation, prototyping, and iteration; and the third day focused on project completion, final presentations, judging, and celebration. Cash prizes totaling $10K were distributed to four winning teams, and 10 Apple watches distributed to runners-up. For many participants, it was their first encounter with HCD practices.
Teams created a litany of projects that addressed mobility products, services, and infrastructure. Notable final projects included a personal wheelchair-docking device designed to support autonomous navigation in rough urban terrains in South America, a responsive bicycle helmet that let cyclists know about hazard issues through location-specific vibrations, and a floating, futuristic mobile assistant that would communicate location information to users.
The following reviews the key elements of a design-a-hack-a-thon based on our first prototype, including recommendations for future iterations.
People. Production partners plan and host design-a-hack-a-thons to better understand a problem space and potential solutions, stimulate public engagement, and/or advance an educational and/or community process around a set of issues. In cases like ours, where the design challenge includes not only products but also services, infrastructures, and policies, intersections between key decision-making partners and designers can support strategic planning that accounts for real-world complexities.
Organizers, who typically work for or closely with the production partners, support event framing, communications, logistics, special needs, and accommodations. They also set the event tone, cultivate best collaborative practices, and respond in real time to emerging event opportunities and challenges. Our organizers were joined by volunteers for setup and day-of-event support.
Teams either register together or assemble with other participants, offering complementary skillsets during opening event sessions. Our student teams represented campuses from across the Boston area (two-thirds undergraduates, one-third graduate students) with diverse skillsets and interests. We recommend teams of three to nine participants that include at least one storyteller, designer, and engineer. Teams might also include local community members and specific groups affected by the design challenge.
Mentors support teams through formal feedback sessions and informal encounters, and by requests for their particular expertise. Mentors help students connect with other mentors, resources, and information; they might host brief sessions on relevant topics or tools. Our mentors included professionals from industry, academia, and policy, and users who deeply understood the challenge at hand. We recommend that mentors attend at least 50 percent of the event to support team progress and community connection.
Judges serve as experts across key topic aspects, such as HCD, policymaking, technical applications, and business. Our judges also served as mentors throughout the event. Special effort should be made to reduce bias, such as clearly specified judging criteria and ensuring that judges connect with all teams. The judge panel should represent a diversity of perspectives and identities.
Community members are users who bring lived experience to the problem space. Users not only bring deep insights throughout the process of finding solutions but also influence solution implementation (e.g., policymakers). Given our broad design challenge, our teams activated existing connections and generated new contacts specific to their project area. Other events might holistically focus on a specific group of users, building relationships in advance. Community members can and should also be invited to join the event as team members, mentors, and judges. The final presentation of projects can be opened as a public event for invitation to the broader community.
Place. Design-a-hack-a-thons involve a range of activities best supported by a mixture of functional spaces. Our event was hubbed in a core convening space large enough to hold all participants. This space included round tables for each team to congregate, AV equipment (e.g., projector screens, portable microphones), and long tables for coffee and food service. Some teams worked in this open space throughout the event, but many used secondary spaces for discussions, prototype development, and displaying works-in-progress. We recommend allocating a space specifically for mentors to convene when not visiting teams, to facilitate mentor interactions and allow teams to locate support. The venue should be proximate to the environments where the design might be implemented, so that teams can observe and engage with people and sites during problem formulation and solution testing.
Process. Pre-event preparations involved advertising the event (Figure 1), content development, participant and mentor recruitment, and communications across email and Slack to generate momentum. Our website briefed participants on event details such as the design challenge theme, event schedule, judging criteria, and registration instructions. We recommend offering participants pre-event HCD resources such as videos describing the design mindset, a question guide to inform their thinking, or even pre-event intensive HCD learning sessions.
|Figure 1. Flyers were designed to advertise the event.|
Our event launched with a kick-off video inspired by the people-centric mobility theme and a series of micro-talks that oriented participants to this design challenge (Figure 2). Organizers and judges then hosted a rapid review of HCD methods. We encouraged teams to deploy four main HCD approaches throughout their process: focus on people, address the correct problem, think about the broader system, and evolve through iteration. Participants were encouraged to focus on people by regularly visiting sites affected by the design challenge (e.g., observing how people crossed a busy intersection close to MIT's campus). Teams were encouraged to conduct user research and create low-fidelity prototypes, undertaking an extensive problem-discovery process before committing to a specific solution (e.g., generating scenarios for communicating between drivers of cars and bikes; Figure 3). Mentors also encouraged teams to identify potential unintended consequences of proposed solutions. An iterative design process defines and redefines problems and solutions, encouraging participants to embrace ambiguity and nonlinearity while moving toward increasingly refined solutions.
|Figure 2. Participants gain contextual information on urban mobility challenges as part of an introductory session on Day 1.|
|Figure 3. A team discusses vehicle communication challenges between cars and bikes as part of an ideation session.|
We organized two formal feedback rounds in addition to offering informal feedback channels throughout the event. On the first day, after time for brainstorming and exploration, teams shared their problem space and a preliminary proposed solution to a group of judges and mentors in five-minute presentations. The second day, we facilitated a one-hour round robin feedback session composed of three 20-minute rounds, in which teams shared in-process solutions and presentations with different mentors. After each of these formal rounds, mentors met to consider challenges shared across teams and direct other mentors toward groups that could benefit from further support. A designated area for mentors enabled teams to find support quickly and gave mentors the opportunity to connect with one another. Slack channels helped teams connect with mentors virtually.
Participants were encouraged to focus on people by regularly visiting sites affected by the design challenge.
The event culminated in two rounds of final presentations, which responded to the judging criteria supplied at the beginning of the event. The five-minute presentations were judged on the team's problem exploration, adaptations to solutions based on prototyping and observations, and the novelty and potential societal impact of their solutions (see example presentation in Figure 4). In the first round, teams were divided into three presentation rooms according to the thematic areas, judged in parallel to select 10 finalists. In the second round, these finalists presented to all participants. A following celebration announced winners and honorable mentions to acknowledge everyone's contributions and support ongoing networking.
|Figure 4. Participants present the features of their team's bicycle helmet prototype during a judging session.|
Culture. Design-a-hack-a-thons create and cultivate the conditions for learning and applying HCD. Empathy, diversity, and collaboration are critical to creating that HCD culture.
Empathy is required for both design-a-hack-a-thon participants interacting with users and for facilitators, mentors, and organizers. Participants must have empathy when conducting observations and interviews, and when building prototypes that uncover people's activities and goals. Understanding the lived experiences, situations, and habits of users is critical to the iterative refinement of needs and solutions. Event organizers, mentors, and judges should also demonstrate empathy for the activities and goals of teams. Participants bring varied levels of familiarity and experience with HCD practices. In just a 48- to 72-hour period of time, it is very difficult to determine needs and to execute a high-fidelity design that addresses those needs. The mentor support and judging criteria should emphasize the process for creating empathetic designs and understanding problem and solution spaces, in contrast to a hackathon's exclusive focus on products and their technical attributes.
Organizers might themselves follow the HCD process in designing the event.
Design-a-hack-a-thons should bring together diverse participants—across interest, talent, expertise, background, and identity. The varied perspectives of mentors and judges might also bring competing advice or suggestions to teams. Team agency, to incorporate or discard advice, should be encouraged. A safe-space policy or culture guidelines can support first-time collaborations, equitable participation, intentional listening, and the expression of vulnerabilities. Our event was open to participants around-the-clock, like traditional hackathons. Events should consider how time structures accommodate different populations; both open time frameworks and clearer boundaries can yield a more diverse and representative participant pool. Pragmatic decisions like schedule intensity, time frames, provisions, explicit expectations, and implicit language often shape who can access the event. Organizers might themselves follow the HCD process in designing the event, working with core user groups or willing participants to test exercises, spaces, and informational materials.
Finally, design-a-hack-a-thons are exciting because the competitive nature of the event is couched within a real-world, collective problem space and focuses more on learning than outcomes. Cultivating a healthy balance between collaborative and competitive motives can support divergence and convergence as complementary aspects of the design process. Following the initial exploration of problem spaces, teams with shared interests may be encouraged to share their works in progress with one another, especially those working on similar challenge spaces. In doing so, teams can challenge one another's ideas, support them, or even merge as appropriate.
Design-a-hack-a-thons offer powerful ways to integrate HCD into the technical hackathon process. Five main criteria can support meaningful reflection and evaluation:
- Ideas. Did the projects identify problems and solutions grounded by interactions with existing and potential users? How desirable and deployable are the concepts and prototypes developed?
- Educational. Did participants learn new tools and skill sets for empathically incorporating users into an iterative, systems-oriented design process? Were these techniques communicated in ways that can be shared beyond event attendees?
- Network building. Did new connections and relationships form across fields? Did the event maximize the opportunity for meaningful and impactful collaborations across disciplines and perspectives?
- Culture. Did the experience of the event generate value, momentum, and reward? Was there alignment between intentions, expectations, and outcome?
- Experience. Did the event provide a quality, human-centered experience for participants and organizers in both body and mind (e.g., provision of healthy food and encouraging rest, facilities for social connection, celebrating process over outcomes)?
Going forward, we see opportunities to evolve this model based on key challenges related to communication, resources, and guidance provided for participants. With respect to communication, we observed the importance of defining key terms to create a shared language among interdisciplinary participants. For example, a prototype in HCD can be as simple as a sketch, but we observed some teams with less HCD familiarity interpret this term to imply higher-fidelity outputs requiring more substantial time to develop, resulting in fewer prototype variations or less time with end users. In terms of resources, we observed that, while teams were provided introductory materials for the HCD process of problem definition on the first day, teams were eager to build and thus often cut short the critical phase of observing and interacting with users. In the future, we plan to expand access to HCD learning tools (e.g., process guides, worksheets, videos, or even pre-session boot camps) prior to the event, so that the orientation session at the event reinforces rather than introduces principles and practices. Finally, we see multiple opportunities to build on the formal and informal feedback structures described here to ensure that teams have the right kind of support at the right time to advance their projects. For example, we have observed in this event and others that it is challenging for participants new to the HCD process to broadly explore a problem space through a divergent process, as the ambiguity associated with this experience can feel uncomfortable, if not at least chaotic. Given this ambiguity, we suspect that teams may find it challenging to ask for and to articulate the kind of help they need. Instead, informal feedback may need to be pushed to teams, with mentors prepared with probes to understand where a team may be stuck, and with event producers developing structures that support responding to teams in real time.
We will continue to build upon the design-a-hack-a-thon model, adapting and shifting it to confront new situations. We hope that this intensive model opens a path for further insights into ways to better integrate HCD processes and technical expertise.
Thanks to Kent Larson, Maggie Church, Phil Tinn, Markus Elkatsha, Michael Lin, and everyone from the MIT Media Lab, as well as Michèle Morris, Colleen Emmenegger, Don Norman, Eric Richards, and participants from the UC San Diego Design Lab for their support in developing and facilitating this design-a-hack-a-thon. We also want to acknowledge the incredible work of our 250 participants. Photos were provided courtesy of MIT Media Lab, City Science.
1. Event page for City Robotics Design-a-Hack-a-thon for People-Centric Mobility; https://www.media.mit.edu/events/city-robotics-hackathon/
2. Dupla Studios. Design swarms: A design process with a twist. Oct. 26, 2016; https://medium.com/@duplastudios/design-swarm-a-design-process-with-a-twist-41f36a93b80f
3. McInnis, B., Xu, X.T., and Dow, S.P. How features of a civic design competition influences the collective understanding of a problem. Proc. of the ACM on Human-Computer Interaction 2, CSCW (Nov. 2018), article 120.
4. Tucker, J. D., Tang, W., Li, H., Liu, C., Fu, R., Tang, S., Cao, B., Wei, C., and Tangthanasup, T.M. Crowdsourcing Designathon: A new model for multisectoral collaboration. BMJ Innovations 4, 2 (2018), 46–50.
5. Vanka, S. Authentic design founder on swarm design. Products That Count. Apr. 22, 2019; https://productsthatcount.com/resources/video/authentic-design-founder-on-swarm-design/
Jennifer Taylor is a postdoctoral scholar with the Design Lab. She cultivates relationships between the university and external partners to support collaborations grounded in community needs. She seeks to approach design with and for people to build products, services, and systems supporting health and well-being. [email protected]
Stephanie Sherman is a Ph.D. candidate working at the intersection of social and speculative design. Her projects, writing, and research convert outmoded sites, systems, and surplus resources into platforms for collaboration and co-production. She works with the Center for Design and Geopolitics and the Design Lab at UC San Diego. [email protected]
©2020 ACM 1072-5520/20/05 $15.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 ACM, Inc.