Crafting Interactive Circuits on Glazed Ceramic Ware

We used a small dining plate to make a resist-blasting swatch (Figure 10). This plate was manufactured in stoneware coated with a white glaze, which results in a distinct color difference between the glaze and underlying ceramic (Figure 4D).

At the center of the plate we arranged a radial pattern of lines. We varied the width of the sandblasted traces on one side; and on the other side, we varied the width of the gap between sandblasted traces. From this experiment, we observed that resist-blasting maintains gaps as small as 0.2mm. However, resist-blasting could only penetrate traces with widths larger than 0.25mm.

We covered the surrounding space with a series of patterns to experiment on. Each pattern was individually sandblasted based on two parameters: distance from the sandblast nozzle and duration of sandblast exposure. We used two types of patterns to highlight different resist-blasting features—a round region to highlight sandblasting spread, and an arcing trace to highlight the sandblasting effect on varying widths. These patterns reveal two important insights about the consistency of resist-blasting: 1) Material removal is most aggressive at the center of the nozzle and tapers off; sandblasting further away from the object results in slower but more consistent material removal. 2) Material removal occurs more rapidly on a thick trace than a thin trace.

We used the same type of plate to make a conductive trace swatch (Figure 11). This plate systematically catalogs conductive traces in terms of depth (shallow vs. deep), trace width (0.5mm to 1.5mm), and type of conductive ink (carbon, nickel, silver). We also demonstrate how a continuous conductive trace might wrap over the rim of the plate from front to back.

This plate reveals the texture of conductive traces and how they are inlaid into the sandblasted tracks. We measured the electrical resistance of each trace and labeled them in the swatch document. The individual measurements highlight the inconsistent electrical resistance between similar traces due to manual painting. However, the overall measurements indicate the consistent effect that trace depth has on electrical resistance—shallow traces exhibit higher resistance, while deeper traces (and therefore a bigger cross section area) exhibit a lower resistance.

Glazed ceramics can be etched with a laser cutter. Although this process is out of the scope of the exploration in this work, we documented laser etched traces on the same plate as a comparison and as a prompt to inspire future work. From this plate, as well as other laser etching experiments, we observed that the laser cutter is less effective at removing material compared to sandblasting—and was not able to completely remove the glaze. Furthermore, while this digital fabrication process removes the need to mask an object, it is only able to work on flat pieces.

We took the same capacitive touch slider design (Figure 15A) and applied it to the side of a porcelain salad bowl (Figure 18A). The recessed traces of the slider on the surface of the bowl offer subtle tactile feedback during interaction. We developed a Processing² visualizer to indicate touch events that occur along this slider. This system detects discrete touch events on each individual region on the slider, as well as sliding gestures by computing the sequence of touch events that occur. We propose that such interfaces could be connected to the smart home network and serve as an alternative means of controlling other home devices during mealtime (such as music or lighting systems).

We instrumented the eating surface of a porcelain side plate with four capacitive sensing traces (Figure 18B). Food-safe porcelain paint was applied on top of these traces to prevent food from coming in direct contact with the conductive trace. These traces are able to sense the proximity of metal utensils in use (e.g. cutting a cake with a dessert fork). We developed a Processing visualizer that renders the capacitive touch readings of each trace. The readings across the four channels vary as a metal utensil moves around the plate during the meal, and these readings can therefore be used to approximate eating activity. We propose that such plates could serve as an interface for monitoring eating activity or for playful mealtime interactions—supporting human-food interaction scenarios previously discussed in HCI research (e.g. [21, 27]).

We applied a heating element to the bottom of a small plate (Figure 21). This plate is able to reach a top surface temperature of 105°C when supplied with 30W of power (12V and 2.5A). Such plates can be used to keep food warm during meal service, and we demonstrate its heating capabilities by cooking a raw chicken egg. We used cork sheets as the placemat material in this application for better heat resistance.

Besides dining applications, we also looked into caring for houseplants (Figure 18C). We applied a moisture sensor to the inner surface of a small plant pot. The moisture sensing electrodes are placed on opposite sides, and each sensing electrode wraps around the outer surface to the base of the pot where it connects to the placemat (Figure 19A,B). We use this sensor to approximate the overall moisture level throughout the soil in the pot by measuring the change in resistance between each electrode. We render the sensor reading on a Processing visualizer through a simple bar that indicates the soil’s moisture level.

While beyond the design goals that we set, the application prototypes—as functioning electronic systems—revealed larger infrastructural needs that we need to address in order for these systems to be properly integrated into the home context. For example, while the acrylic placemats were a clever way to connect the vessel to a microcontroller, vessels had to be carefully placed so that their conductive pads align with the pins on the placemat. In addition, we were using a basic microcontroller (Arduino Uno R3) to parse the signals from the ceramic sensors, and sending these signals to a laptop computer for network communication and visualization. While this served as a convenient setup for a research prototype—it is certainly too cumbersome for long term interactions with these interactive objects as research products [31]. These gaps point us toward practicalities to consider for future work. We are inspired by platforms like Lilypad [5], Chibitronics [2], and Bare Conductive [1] that provide extrinsic hardware support for incorporating electronics into physical crafts surrounding textiles and paper. Similarly, we see an opportunity to develop specialized hardware components such as networked microcontroller breakouts and electrical connectors that work with the interactive ceramic objects built with our approach. We also see an opportunity to extend our collaboration to interior designers, builders, and electricians to explore how ceramic interfaces and the electronics infrastructure they require can be integrated into the construction of our living spaces. Such components and processes will enable us to build higher fidelity systems that facilitate long term and real life engagements with these artifacts.

Our approach supported the construction of ceramic artifacts that can withstand everyday use. Notably, the recessed nature of the conductive traces meant that they could be easily cleaned without damaging the circuit (Figure 22). Beyond these short term engagements with the artifacts, we also considered the longer life cycle of these electronic ceramic interfaces. In this regard, we were challenged with working with things that were inherently prone to breakage; and indeed, we broke a plate during the course of our research that we spent a good amount of time working on. This led us to reflect on issues surrounding the sustainability and end-of-life for computational devices [4, 43, 53]. We were reminded of the Morse Things project and how it commemorated broken prototypes through kintsugi repair [32]. Breaking our prototype provokes us to consider how we might repair, restore, or renew the interactive ceramic devices that are built from our approach. Coincidentally, these are themes that Hans Tan addresses in his work as well, albeit in the field of decorative arts instead of HCI. We see an opportunity to extend our work beyond crafting tangible interfaces with resist-blasting—and consider how adjacent repairing practices might participate in upkeeping interactive ceramic artifacts.

In this research, we made a conscious effort to keep the processes and tools we explored as close to the ceramics workshop as possible. Our close collaboration with Hans Tan was essential to this work. Collaborating with a ceramics craftsperson was a way for us to have a proxy in a material practice we were not yet familiar with. We, then, were a proxy for the material practice of physical computing; understanding what would be required for achieving various functional effects. As discussed by Frankjaer & Dalsgaard [12], we found ourselves moving between “localized points of interests” between the “material landscape” of glazed ceramic ware and physical computing—seeking to open up paths between Hans Tan’s practice in ceramics and our practice in making interactive systems.

This led us to new tools and processes that we had not considered before for enabling physical computing on glazed ceramic ware. As a notable example: we did not expect conductive paint to dry so rapidly, which disrupted our initial assumption of wiping excess ink off as a liquid. In closely assessing and experimenting with the dried ink on the glazed ceramic surface, we discovered that we could scrape it off cleanly. The approach we were developing took an alternative route—and one that was more viable and relevant to the materials present. When we made a trip to the crafts supply store to source for the appropriate scraper, we were greeted with a whole suite of scraping implements that ceramic crafters already use as part of their process.

This partnership helps us understand the HCI practice we took in this work as one that centered the established modes of ceramic craft in the production of new interactive possibilities. This practice both illuminates a new approach to circuits, as well as the relevance of “old” craft to present day innovation. We hope that the productive partnership we forged between ceramic crafts and HCI will contribute to how we (as a research community) consider incorporating computational technology into everyday life such that “value flows in all directions” between new and existing object systems [34].

The process we took is inspired by prior work in HCI where researchers collaborate with craftspeople for the production of interactive artifacts; such as in leather-working [47] and ceramics [9]. Our research further the forms of such collaborations by highlighting the potential of partnering with an individual and their unique craft practice for creating interactive systems; alongside learning and extracting general techniques used in a particular realm of craft.

The work we present in this paper goes beyond simply shaping conductive paint on a ceramic surface. We investigated the double-layer composition that glazed ceramic objects offer, and the affordances of this physical structure for functional electronics. A large part of our exploration revolved around the specialized resist-blasting technique developed by Hans Tan. We sought to “align and integrate” [12] this crafting technique with the physical computing principles that we were familiar with to build functional interactive systems. Exploring physical computing through the lens of the structure of glazed ceramic objects and a subtractive fabrication process revealed many new affordances for electronics.

Most crucial to our approach—resist-blasting enabled intricate traces to be etched onto the three dimensional surface of the ceramic artifacts. It exposes the rough ceramic body as a suitable substrate for receiving conductive paints, and preserves the hard and smooth outer glaze to facilitate scraping off excess paint. Besides circuit construction, the structure of glazed ceramic also facilitated the operation of sensors and transducers. The waterproof glaze supports resistive moisture sensing by channeling water effectively to the electrodes (Figure 15C), while the thermal transmissivity of ceramic supports effective transfer of heat generated by the heating element through the walls of the object (Figure 21).

In collaborating with Hans Tan as an individual craft practitioner, we also found ourselves contributing back to his unique process of resist-blasting. To design the temperature sensing tile (Figure 15B), we developed a computational design algorithm to pack a long trace inside a small circular area. We were initially focused on maximizing the length of a trace to improve the sensor’s response to temperature change. As we worked on this tile, we noticed that these computationally generated traces also facilitated other steps in the making process. The pattern, despite its complexity, was easy to cut with the Cricut machine as it consists of a single unbroken path. The branching coral-like traces provided the vinyl mask strong adhesion on the glazed surface to withstand sandblasting (Figure 24A). These same branches also facilitated efficient scraping of excess conductive ink (Figure 24B).

Designing and making the temperature sensor serendipitously revealed new insight into the masking process for resist-blasting that our craft collaborator had not considered before, broadening the range of patterns that can be created with the technique. This insight stemmed from a computational design process that generated patterns that would have been challenging for us (including the craftsperson) to conceive manually.

HCI researchers proposed that digital technology can broaden the expression of existing craft [16]. For us, the temperature sensor reveals one form of such productive partnership between traditional craft and digital design. It simultaneously demonstrates the role of computational design and fabrication in informing and widening a traditional craft technique, while also revealing the role of traditional craft plays as a context that gives real world meaning to explore these digital practices. We see more potential in this intersection of digital and traditional processes for crafting interactive systems, and aim to investigate other computational design algorithms and digital fabrication processes along with ceramic craft practices in future work.