Designing For Human-Centered Automation: A Co-Design Study with Fabrication Professionals Nur Yildirim Introduction Carnegie Mellon University Intelligent systems increasingly collaborate with people, Human-Computer Interaction automating a variety of work tasks and tasks of daily liv- Institute, Pittsburgh, PA, USA ing. One constant challenge when designing for human- yildirim@cmu.edu centered semi-automated systems is to negotiate the bal- ance between human agency and machine autonomy. In this position paper, I share a co-design study that investi- gated the current and desired experiences of professionals in micro- and small-scale manufacturing settings. Through this exploratory research, I aim to highlight (1) op- portunities and pitfalls for semi-automated intelligent sys- tems in the context of digital fabrication and (2) the role of design research in exploring the boundaries between ma- chine autonomy and human agency. I identify and discuss the dimensions for designing human-centered automation that enhances professionals’ felt experience of work with- out threatening their autonomy. By sharing these insights, I hope to start a discussion about the complexities of situ- ating machine intelligence and automation within specific contexts. Case Study My ongoing research investigates how to enhance the user experience of fabrication systems through machine au- ________________________________________________________ tonomy and intelligence. Digital fabrication tools (DFTs, Workshop proceedings Automation Experience across Domains e.g., 3D printers, laser cutters, CNC routers) have trans- In conjunction with CHI'20, April 26th, 2020, Honolulu, HI, USA Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). Website: http://everyday-automation.tech-experience.at formed micro- and small-scale manufacturing and impacted Throughout the sessions, I explored the boundaries be- the practices of professional craftspeople. These semi- tween machine autonomy and user agency by using probes automated tools give users superhuman abilities by mak- such as “If your tool was intelligent enough to do anything, ing the fabrication process faster, more precise and more which part would you keep to yourself in the process of repeatable. However, with the arrival of DFTs, there has making?” These discussions were essential in discovering been some concern in the HCI community about a loss of which aspects of a task feel mundane and which aspects agency, craft skill, creativity, and pleasure in making [2, 6]. are critical to people’s perception of agency. The workshop outcomes revealed several dimensions through which au- The presented case study is an exploration of the current tomation and machine intelligence can enhance profes- needs and future desires of fabrication professionals (e.g., sionals’ felt experience of work without threatening their custom manufacturers, fabrication specialists, shop stew- autonomy. ards) with an eye for opportunities and pitfalls for automa- tion. Professionals in small-scale production is an interest- Desire for Increased Automation and Intelligence ing group to explore the intersection of creative practice, Fabrication professionals expressed a desire for more au- control, and machine intelligence as they have a high-level tomation and intelligence in the process of digital fabrica- of agency over their work (unlike other users in manufac- tion. Their concern for a loss of agency was unexpect- turing, e.g., factory workers [5]). The goal of this work was edly less, given the literature and previous HCI studies. to surface insights on how these tools might become more Almost all participants wanted fabrication systems to au- intelligent and capable to automate mundane work in ways tomate tasks such as machine setup, material registration, that do not negatively impact the human experience. calibration, and maintenance. When asked to envision fu- ture DFTs, they described self-aware fabrication systems Co-Design Workshops that can sense and act upon their performance and the “From a business standpoint, To investigate possible futures for machine autonomy and workspace. They wanted intelligent DFTs that have aware- the less I have to do personally its impact on user agency and control, I conducted a co- ness of users’ high-level goals and that can self-adapt its the more money I make. On design study with 23 fabrication professionals. I asked, plans based on results to achieve these goals. my own personal projects, I’d “What if DFTs gained more intelligence to the point that still rather just push a button they can be active collaborators in the fabrication process? Boundaries Between Autonomy and Agency and have the part come out.” How would DFTs with increased intelligence and automa- Professional users’ concern for control and agency often (P19) tion impact users’ felt experience of work?” As a construc- centered on the inability to customize DFTs. They stressed tive activity, the co-design workshops enabled the partic- that they “should be able to tweak and take manual con- ipants to actively participate in envisioning the preferred trol over the settings” if they didn’t like the auto-generated futures instead of only focusing on breakdowns in current ones. Some participants voiced concerns around paternal- fabrication systems. istic automation that impose limitations on users to provide “ease of use”. They described a spectrum of automation The co-design activities included “tell” and “make” [1], where a variety of tasks are automated for error prevention where participants told stories about their current expe- and efficiency, yet they can remain in control when needed. riences and brainstormed future interactions with DFTs. Negotiation of Time-Quality-Cost Trade-offs 1. What are the dimensions of human-centered automation? Professional users’ desires, as well as their discussion The findings of this fieldwork revealed productivity-oriented around future intelligent DFTs, revealed a lack of support dimensions (e.g., error prevention, efficiency, self-maintenance, for making trade-offs between time, cost, and quality in cur- safety) as well as experience-oriented dimensions (e.g., rent systems. Participants shared that they struggled while personalization, decision support, skill development, re- making these trade-off decisions, which are emergent in source curation) for automation technologies. What other “[System could say] ‘Do you the situation. They wanted DFTs to have awareness over dimensions might exist for human-centered automation? realize that at this level you’re the quality of the outcomes (e.g., surface roughness, di- How might we, the HCI research community, build frame- getting this quality, do you need mensional accuracy) and the controls to achieve those out- works for automation that go beyond the paradigm of pro- that quality? You can have this comes so that systems could help users to arrive at a set of ductivity? How might these frameworks generalize across or that.’ (P11) trade-offs. domains and how might they situate into specific contexts and user groups? Intelligent Shop Assistant Rather Than Collaborator Fabrication professionals desired DFTs that can leverage 2. How might intelligent semi-automated systems negoti- their machine capabilities to give users superhuman abili- ate the quality of work? ties. They wanted systems to log and recall settings, time, Decision support for making time-quality-cost trade-offs was “[I should to be able to say] This material used, and the outcome as an aid in documentation an emergent need that illustrated how we might design for is what I consider a good cut, and self-reflection. They perceived future DFTs as intelli- an interplay of users and intelligent systems. How does an because I use this laser totally gent shop assistants that actively learn how their users like automated system gain knowledge about its own capabili- different from everyone else.” to work, curate settings, and personalize their operations ties and the quality it is producing in order to negotiate such (P23) towards their users’ tastes. trade-offs? How do the abstract notion of quality and the subjective human judgment connect to the measured per- Prompts for Workshop Discussion formance of a system? Through the above case study, I draw attention to the chal- lenges and complexities of situating machine intelligence in 3. How to develop a design process for blending human contexts where creative practice and automation overlap. agency and machine autonomy? Previous HCI studies indicate that the level of desired au- Can we build on existing design research and interaction tomation may vary between different groups of users, even design methods to develop a design process for human- within the same domain or context [4, 3]. The insights from centered automation? How can we explore, design, proto- this case study can serve as a point of reference for other type and evaluate automation experiences before investing HCI researchers to identify and account for shared and con- in building these systems? flicting desires of different user groups and stakeholders. The following reflections aim to start a discussion: REFERENCES [1] Eva Brandt, Thomas Binder, and Elizabeth B-N Sanders. 2012. Tools and techniques: ways to engage telling, making and enacting. In Routledge Designing For Human-Centered Automation AutomationXP’20, April 26, 2020, Honolulu, HI, USA international handbook of participatory design. Andreas Reiter, and Pernille Bjorn. 2016. Fabrication Routledge, 165–201. & HCI: Hobbyist Making, Industrial Production, and [2] Amy Cheatle and Steven J. Jackson. 2015. Digital Beyond. In Proceedings of the 2016 CHI Conference Entanglements: Craft, Computation and Collaboration Extended Abstracts on Human Factors in Computing in Fine Art Furniture Production. In Proceedings of the Systems (CHI EA ’16). ACM, New York, NY, USA, 18th ACM Conference on Computer Supported 3550–3557. DOI: Cooperative Work & Social Computing (CSCW ’15). http://dx.doi.org/10.1145/2851581.2856491 ACM, New York, NY, USA, 958–968. DOI: [5] Daniela Wurhofer, Thomas Meneweger, Verena http://dx.doi.org/10.1145/2675133.2675291 Fuchsberger, and Manfred Tscheligi. 2018. Reflections [3] Peter Fröhlich, Matthias Baldauf, Thomas Meneweger, on Operators’ and Maintenance Engineers’ Ingrid Erickson, Manfred Tscheligi, Thomas Gable, Experiences of Smart Factories. In Proceedings of the Boris de Ruyter, and Fabio Paternò. 2019. Everyday 2018 ACM Conference on Supporting Groupwork Automation Experience: Non-Expert Users (GROUP ’18). ACM, New York, NY, USA, 284–296. Encountering Ubiquitous Automated Systems. In DOI:http://dx.doi.org/10.1145/3148330.3148349 Extended Abstracts of the 2019 CHI Conference on [6] Amit Zoran, Roy Shilkrot, Suranga Nanyakkara, and Human Factors in Computing Systems (CHI EA ’19). Joseph Paradiso. 2014. The Hybrid Artisans: A Case Association for Computing Machinery, New York, NY, Study in Smart Tools. ACM Trans. Comput.-Hum. USA, Article Paper W25, 8 pages. DOI: Interact. 21, 3, Article 15 (June 2014), 29 pages. DOI: http://dx.doi.org/10.1145/3290607.3299013 http://dx.doi.org/10.1145/2617570 [4] Verena Fuchsberger, Martin Murer, Manfred Tscheligi, Silvia Lindtner, Shaowen Bardzell, Jeffrey Bardzell,