Sketching is a way of exploring early concepts through the act of externalization in a suitable material with the aid of a suitable tool. One could use paper and sketch with a pencil or go digital and sketch with code. What is appropriate to choose depends on the situation and on the skillset of the person who is going to sketch. When sketching is done successfully, the externalization can “speak back” and thus engage the sketcher, and others, in a conversation leading to a better understanding of the sketched concepts as well as new concept ideas. This is thoroughly documented in literature – a typical example would be an architect sketching a site plan on a flat piece of paper and being able to read into the possible movements in the third-dimensional space. Sketching generally works like this in familiar, that is, idiomatic situations for experienced sketchers. In unfamiliar or non-idiomatic situations, existing sketching tools are inadequate for expressing and exploring early concepts. For novice sketchers, with limited sketching literacy, even attempting to sketch in an idiomatic situation can be challenging.

Through three cases, I design for concept exploration by enabling sketching to understand how this can be done in new situations. The first case deals with expert sketchers exploring non-idiomatic situations: professional creatives working with fulldome format for visual communication. The second case deals with novice sketchers exploring non-idiomatic situations: design students working with virtual reality. The third case deals with novice sketchers exploring idiomatic situations: air traffic controllers working with finding alternative routes for aircraft in the airspace with automation support.

I take a constructive design approach by making design examples and reflecting during and after the process. With the help of the design examples, I engage domain experts through participatory co-design workshops and elicit insights in order to inspire further design work. What I learn through this dynamic making-workshopping-and-reflecting process forms the foundation of the knowledge contribution. It is presented here as three design tactics on how sketching could be like to support concept exploration: 1) be responsive, 2) emulate salient material properties, and 3) be lightweight.

Att skissa är ett sätt att utforska tidiga koncept. Det görs vanligtvis genom att externalisera koncepten i ett lämpligt material med hjälp av ett lämpligt verktyg. Det går likväl att skissa med papper och penna som att skissa digitalt med hjälp av kod. Vilket material eller verktyg som är att föredra beror på situationen och skickligheten hos personen som ska skissa. När skissning sker framgångsrikt kan externaliseringen “prata tillbaka” och på så sätt engagera skissaren, och andra, i en konversation som leder både till bättre förståelse för koncepten och även till nya konceptförslag. Det här finns utförligt beskrivet i litteraturen – ett typexempel är när en arkitekt skissar en byggnadskonstruktion på ett plant papper och utifrån det kan se hur möjliga rörelser kan te sig i en tredimensionell rymd. Skissning funkar generellt sett på det här sättet i alla fall i välbekanta, idiomatiska situationer för erfarna skissare. Däremot i obekanta, icke-idiomatiska situationer, är befintliga skissningsverktyg otillräckliga för att kunna användas till att uttrycka och utforska tidiga koncept. För novisa skissare, med begränsad skiss-kunnighet, är det dessutom utmanande att skissa även i en idiomatisk situation.

I tre fallstudier, designar jag för konceptutforskning för att förstå hur det kan ske i nya situationer genom att möjliggöra för skissning. Den första fallstudien handlar om erfarna skissare som utforskar icke-idiomatiska situationer: professionella kreatörer som jobbar med domformat för visuell kommunikation. Den andra fallstudien handlar om novisa skissare som utforskar icke-idiomatiska situationer: designstudenter som jobbar med virtuell verklighet. Den tredje fallstudien handlar om novisa skissare som utforskar idiomatiska situationer: flygledare som jobbar med att hitta alternativa rutter för flygmaskiner i luften med automationsstöd.

Jag tar en konstruktiv designansats genom ett skapande av designexempel och reflektion under och efter processen. Med hjälp av designexemplen engagerar jag domänexperter i deltagande co-designworkshops och fram insikter som inspirerar fortsatt designarbete. Det jag lär mig under den här dynamiska processen av skapande, workshopande och reflekterande, formar grunden till kunskapsbidraget. Bidraget presenteras här som tre designtaktiker om hur skissning kan vara för att stödja konceptutforskande: 1) vara responsiv, 2) emulera centrala materialegenskaper, och 3) vara lättviktig.

Recent advancements in machine learning (ML) have led to a dramatic increase in AI capabilities for medical diagnostic tasks. Despite technical advances, developers of predictive AI models struggle to integrate their work into routine clinical workflows. Inefficient human-AI interactions, poor sociotechnical fit and a lack of interactive strategies for dealing with the imperfect nature of predictions are known factors contributing to this lack of adoption.

User-centred design methods are typically aimed at discovering and realising desirable qualities in use, pragmatically oriented around finding solutions despite the limitations of material- and human resources. However, existing methods often rely on designers possessing knowledge of suitable interactive metaphors and idioms, as well as skills in evaluating ideas through low-fidelity prototyping and rapid iteration methods—all of which are challenged by the data-driven nature of machine learning and the unpredictable outputs from AI models.

Using a constructive design research approach, my work explores how we might design systems with AI components that aid clinical decision-making in a human-centred and iterative fashion. Findings are derived from experiments and experiences from four exploratory projects conducted in collaboration with professional physicians, all aiming to probe this design space by producing novel interactive systems for or with ML components.

Contributions include identifying practical and theoretical design challenges, suggesting novel interaction strategies for human-AI collaboration, framing ML competence for designers and presenting empirical descriptions of conducted design processes. Specifically, this compilation thesis contains three works that address effective human-machine teaching and two works that address the challenge of designing interactions that afford successful decision-making despite the uncertainty and imperfections inherent in machine predictions.

Finally, two works directly address design-researchers working with ML, arguing for a systematic approach to increase the repertoire available for theoretical annotation and understanding of the properties of ML as a designerly material.