Mannequette: Understanding and Enabling Collaboration and Creativity on Avant-Garde Fashion-Tech Runways

Drawing upon multiple disciplines, avant-garde fashion-tech teams push the boundaries between fashion and technology. Many are well trained in envisioning aesthetic qualities of garments, but few have formal training on designing and fabricating technologies themselves. We introduce Mannequette, a prototyping tool for fashion-tech garments that enables teams to experiment with interactive technologies at early stages of their design processes. Mannequette provides an abstraction of light-based outputs and sensor-based inputs for garments through a DJ mixer-like interface that allows for dynamic changes and recording/playback of visual effects. The base of Mannequette can also be incorporated into the final garment, where it is then connected to the final components. We conducted an 8-week deployment study with eight design teams who created new garments for a runway show. Our results revealed Mannequette allowed teams to repeatedly consider new design and technical options early in their creative processes, and to communicate more effectively across disciplinary backgrounds.


INTRODUCTION
Avant-garde describes artistic, intellectual and cultural movements that are characterized by experimental, radical and unorthodox approaches [12]. The capacity in which something goes beyond the boundaries of convention through experimentation and proposes a new way in the creative arts -e.g. art, dance, fashion -is what makes it avant-garde [12]. In the fashion industry, avant-garde has impacted the way designers think and create work, especially around the form, shape and volume of outfits worn today. The fashion industry, along with avant-garde, influences culture and global trends [32], drives social and economic change, and is a fundamental component of everyday life. Today, both are rapidly being transformed with modern technological innovations -such as factory robots which handle garment construction and manufacturing processes, algorithms which predict trends in style and even VR mirrors and installations [20,33].
The expression of fashion (i.e. its communication mediums) can be divided into two categories: the street (or consumer market) and the runway (or catwalk) [32]. Avant-garde runways have incorporated numerous technologies (e.g. robotics, sensors, fiber optic fabrics) for several years, with designers frequently collaborating with technologists, and other disciplines, forming teams that create garments that are infused with technology. Throughout the collaboration and 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. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Permissions@acm.org. creation processes for a team creating avant-garde fashiontech garment -from sketch to runway -designers, technologists and others, often draw upon a number of different disciplines, such as fabrication, electronics, and programming (among many others). Several tools can also be used to help novices in learning these disciplines that are essential for prototyping and creating avant-garde fashiontech garments [1,38]. However, these tools also present numerous challenges within creative and collaborative processes of multi-disciplinary teams that create avant-garde fashion-tech garments which appear on runways. Within these teams, designers, technologists (and others) often operate with unique skill sets, knowledge, and vocabularies making collaboration sometimes difficult. Thus, the problem is that these collaborators do not have a common medium to express creative ideas, experiment with sometimes complex technology (e.g. sensors, lights, augmented reality) and communicate with one another about challenges and limitations, particularly in the early stages of constructing an avant-garde fashion-tech garment.
Our approach is to support facilitated communication and prototyping between team members (e.g. fashion designers, technologists) by the introduction of the Mannequette (Figure 1). Mannequette is a modular, miniaturized mannequin system, designed to help cross-disciplinary teams in prototyping and experimenting with technologyspecifically lights and sensors -from the onset of the garment construction process. With Mannequette, a team can engage in low-fidelity prototyping of sensors and lighting patterns using a simple DJ-mixer like control panel, without the immediate reliance of a technologist or technical skills. A technologist can also adjust the sensors and light patterns created by a fashion designer (or other team members) with the control panel through programming, for demonstrating more complex interactions and effects. Thus, Mannequette serves as a communication medium between team members, as it can be used to create and illustrate temporal effects that may otherwise be difficult to communicate.

RELATED WORK
In the context of designing and creating tools to facilitate and support the process of constructing of avant-garde (and fashion-tech) garments, we drew upon two primary research areas to inform our design: (i) fashion technology industry and research, (ii) programming tools and construction kits for wearables. We also position our contributions within the space of wearables, fashion tech and existing tools.

Fashion and Technology
The fusion of fashion and technology has transformed garment creation and fabrication processes [8,27], now making it a more critical component and influencer of the consumer wearables industry, and vice versa. One major area in fashion seeing this influence is e-textiles, which incorporate electronics that are woven into the fabric and has a number of applications [13,16,18,19,27,28]. Technologies such as 3D printing sensors, actuators and other components have also begun being incorporated into clothing [26,28,30,31] and even on the skin [11,21,22,35,36]. In this context however, several fundamental challenges still exist, such as the need for technology to be lightweight or finding sufficient and long-lasting sources of power. These challenges and the ability to solve them creatively, rely on a cross pollination of skills between design, engineering and other disciplines. In the world of avant-garde and fashion tech, the experiences and lessons learned in creating garments can lead to or inspire more innovative wearable solutions to these challenges.
In the broader context of HCI research, prior work has examined several application areas in fashion using technology, such as in-store environments [7], while other researchers have focused on the relationships between fashion and clothing [15], understanding the design of electronics with fashion lenses [14] and design principles from fashion itself [24]. Pan and Stolterman described scenarios for HCI driven by fashion from different perspectives, such external fashion concepts influencing the field, as well as fashion concepts that go in and out of fashion within the field itself [25]. Okerlund  as a lens to study empowering a campus Maker community [23]. We build upon the approach of using a fashion show as the context for conducting research, as well as the work of Pan and Stolterman, but focus on the underexplored avantgarde fashion-tech area and its cross disciplinary collaborations.

Toolkits for Designing Wearables
A large number of electronics and wearable construction kits, such as the Adafruit Flora [1,39], the LilyPad [5,6] ,the BBC Micro:bit [2] and other Arduino-based electronics kits [40] have opened up a new space for children and adult hobbyists to begin exploring and building new kinds of wearables and integrate electronics into clothing. However, many of these toolkits and electronics still have high barriers of entry, requiring knowledge in areas such as electronic circuits, an understanding of lower-level digital and analog I/O (especially if sensors for interactivity are used) as well as programming. Several toolkits have been designed to overcome these issues especially with regards to programming and electronics [40]. However, only a limited subset is aimed towards avant-garde fashion-tech, and even less are appropriate for the constraints that come with designing for and building garments that are worn a runway. For example, using conductive threads to link components, a common approach for several toolkits (e.g. [5]) is not ideal due to: (i) movement of garments on a runway; (ii) the wear and tear of a garment over its lifetime as it travels to different runway shows, while also being worn by different models.
One approach to address the difficulties in programming, electronics and runway constraints is to consider tangible and modular approaches. MakerWear is an example of using a tangible and modular approach for wearables [17], similar to littleBits and others [2,3,41,42]. While MakerWear is targeted towards building wearables for children and demonstrated value in enabling children to begin working with basic programming and sensors for wearables, its focus and design is not targeted towards fashion designers, technologists (i.e. avant-garde fashion-tech teams) or the runway environment. We build upon this and prior work in tangible construction kits [1,6,17,42], where the construction experience (both user input and output) for Mannequette is tangible, serving as a means of communication, creative expression and experimentation without many barriers at the onset of constructing an avant-garde fashion-tech garment.

DESIGN CONTEXT
To ground our research within the context of real-life practices of avant-garde fashion-tech runways, we collaborated with a fashion-tech collective (MakeFashion [43]) that produces multiple international avant-garde runway shows per year. We built our relationship with MakeFashion over a 2.5-year period through 7 separate runway shows in Canada, USA, Germany, Ireland and China. Each show contained 8-15 design teams, where each team created up to three avant-garde garments per runway show. Design teams were typically comprised of at least one self-identified fashion designer, and one self-titled technologist. The construction of teams (and their selfidentification with titles) are specific to the context of MakeFashion; thus, the bifurcation of technologists and fashion designers does not necessarily reflect the broader community of avant-garde fashion-tech, many of whom have cross-disciplinary skills. For the purpose of this paper, we use these labels to describe and reflect the people we studied. We return to the issue of generalizing beyond this community later in our discussion.
To understand the challenges and process of avant-garde fashion-tech and runways, we embedded ourselves in three key roles for runway shows: first as a technologist, then as a fashion designer, and finally as a producer and director. Thus, the lead author used a participant observer process in different roles within this organization to understand the practical problems faced by members of each group. As a trained computer scientist, the lead author began involvement in the organization as a technologist within a design team. After three runway shows in this role, the lead author then organized a new design team, and played the role of a fashion designer on the team for another three runway shows. In total, the involvement as a technologist and designer resulted in 14 garments, and interaction with 30 other designers and technologists. Since then, the lead author has been a producer and technical director for MakeFashion, managing and meeting with teams over the course of 4 months prior to the runway show. As a consequence, the lead author has worked with over 20 teams, producing over 40 avant-garde fashion-tech garments.
Next, within the context of MakeFashion, we describe in detail, our observations of common practices in avant-garde fashion-tech garment creation.

Observed Practices of Avant-garde Fashion-Tech
Based on our observations and experiences over the 2.5-year period, we saw three distinct phases based on increasing technical complexity and involvement of a technologist. This process is briefly summarized in Figure 2, where each of the lettered phases circumscribes different processes that are commonly happening within a cross-disciplinary team that typically consists of at least one self-identified fashion designer and one self-identified technologist.

Phase A: Inspiration
Before creating a garment, a team (typically the fashion designer) explores different sources of inspiration-Instagram, Pinterest, magazines, pattern drafting books, music-with the goal of creating a story that describes the garment on a runway. Story plays a critical role for avantgarde garments on a runway, as it is ultimately translated into the design of the garment, the technology used to support it, the music chosen for the runway, as well as how the model walks and poses on the runway. Furthermore, while fashion has a long history and set of materials to draw upon for inspiration, avant-garde fashion-tech is a relatively newer space, thus fashion designers and teams are more limited in examples and materials they can draw upon for inspiration.
Once a concept has been settled upon, low-fidelity experimentation is used to explore possibilities for implementing the idea. Tangible and physical techniques are also used to prototype different elements of a garment, such as testing different fabrics and materials to understand how well lights diffuse them, color combination of fabrics and the draping of fabrics on a mannequin. Fashion designers typically create sketches for the concept of the garment, which is occasionally combined with a limited description of what is expected from technology.
This phase is typically characterized by limited involvement from a technologist, and any progress is revisited once the technologist becomes more deeply involved. In this phase, limited technical engagement creates two challenges: (i), since sketches and creative ideas are initially developed without a clear understanding of technical limitations, ideas often need to be scaled back drastically later in the design or construction process; (ii), since designers are rarely wellversed in how or what technologies could be deployed, their ability to communicate is limited, meaning communication opportunities are lost.

Phase B: Experimentation
Once a garment (story and concept) has been defined, a fashion designer and a technologist begin using high-fidelity techniques and technical "prototypes" are also built -often independently. Teams typically face challenges when trying to integrate these prototypes, owing to situations where intentions of the design were unclear, or limitations of the technology were not clearly communicated earlier.
Here, fashion designers develop multiple iterations and prototypes until they are satisfied with the outcome of their pattern and chosen fabrics. They begin with sketches, and draft with patterns appropriate to the design. Those with access to fabrication skills (i.e. laser cutting and 3D printing) use them to speed up or augment different parts of their pattern drafting process (e.g. using laser cutting to quickly prototype modifications to a pattern).
In parallel, technologists' experiment with different electronics and components (e.g. motors, haptics, lights, sensors), create programs, and occasionally design new electronics entirely for a garment. For example, with a light strip, they may explore different light patterns, their timing or colors. Typically, a technologist's role does not have a large contribution to the design itself, and thus it is extremely common for the technologist to misunderstand an idea articulated by the fashion designer. Thus, the technologist might develop the technical prototype in ways that are unexpected or unwanted by fashion designers.
Once the designer and technologist are respectively satisfied with their choices, they begin garment construction and integration. Teams run into significant problems in this step, many of which are due to poor communication practices, and some due to practical problems of deployment. Typically, the technology integration (i.e. microcontrollers, wiring, sensors) occurs after the garment has been constructed, which has negative consequences. For example, many teams fail to consider where sensors should be placed (or sewn into the fabric), or where (or how many) batteries must be placed on the body. This means that garments require additional work, such as constructing pockets for batteries or sensors, potentially designing new 3D printed and sewable casings for microcontrollers or building housing for extremely messy wiring. This problem is even more difficult for novice teams of fashion designers and technologists, who have limited experience designing for wearability on the runwaysometimes, their early fixes to the garment are inadequate (e.g. gluing LED pixels vs. building an LED layer), meaning they make harmful compromises to the garment itself.

Phase C: Runway
An entirely new set of challenges present themselves when the garment is worn by a model in preparation for a runway show. While the garment may be functional, teams can encounter problems during rehearsals. Here, practical challenges of wearability present themselves again-for instance, was the fabric chosen for the design flexible enough to accommodate the model's walking movement or poses, is there enough slack in the wiring for electrical signals to pass through during choreographed movement (as may be required by the story), and does sweat from the model damage or cause problems for the technology used?
We frequently observed that when a garment was involved in dance or other choreography, electronics were at risk. These issues were typically uncovered during rehearsals. For example, wires may have broken or stretched underneath a casing, or the movement programmed into accelerometer to cause an effect, is too sensitive for the walk of a model. This meant that designs adjustments and technical adjustments needed to be debugged and made on the fly. This was especially challenging if the design of the garment did not account for the ability to make necessary changes when something breaks down, such as a wire or sensor. For novice teams of fashion designers and technologists, these issues were magnified because of the late integration of technology, with consequences now including the removability and reparability of the garment. Overall, many teams also do not have prior experience in understanding potential runway show issues for avant-garde fashion-tech and their solutions.
Following a runway show, a garment typically requires repairs and adjustments which might again require taking apart the piece. Furthermore, garments commonly travel between shows and are worn by different models, presenting additional challenges. For instance, garments often need to be put on a model in a certain way or in a certain order (e.g. some pieces have multiple layers) and transferring this knowledge to others who do have not the necessary background or experience with garments incorporating technology is difficult. Additionally, technical knowledge of avant-garde fashion-tech garments is not well maintained or documented. This becomes a problem when a garment requires repairs, and its creators have not travelled with it. This also creates several issues around communication between different groups of technical and fashion expertise and can substantially hinder the success of the garment in the highly-competitive avant-garde fashion world.

The Challenges of Avant-Garde Fashion-Tech
Traditional fashion runway shows are messy and hectic by their very nature and by combining avant-garde culture with technology to the mix, a variety of challenges have been created. We distill many of these challenges we observed previously into three themes. Balancing Wearability and Troubleshooting. Building garments for runway conditions is difficult, particularly when garments are also worn by different models in its lifetime. This makes designing for wearability extremely important when considering the types of sensors, fabrics, lights and batteries that can be used in constructing a garment. However, a balance needs to be struck between the wearability and the ability to fix the garment, as they break down and need repair overtime. Because technology is incorporated later in the design process, these considerations are not thoroughly accounted for by the designer or technologist in the early stages, resulting in garments that are wearable but not repairable (or vice versa).

MANNEQUETTE
Currently, wearable kits support a wide range of sensors and outputs, but they do not necessarily support or properly consider the range of described activities in avant-garde fashion-tech, nor the audience: multidisciplinary teams consisting of fashion designers, technologists and others. We envisioned designing a tool built around the processes we observed while also building upon prior research [6,9,10,17]. Thus, we designed a miniature mannequin-based tool called Mannequette, with the name referring to mannequins (of all sizes) that fashion designers build around, combined with a marionette which is controllable by a puppeteer. Thus, Mannequette is a mannequin with a controllable piece.
Mannequette allows teams (especially fashion designers) to test, experiment and quickly iterate upon light patterns, interactions and fabric techniques (e.g. diffusion testing with light or dark fabrics) in a rapid and ad-hoc manner without technical knowledge, at the onset of the construction process. Certainly, the space of availability of technologies for creative expression in avant-garde fashion-tech is large and while examples of avant-garde fashion-tech garments do exist using different methods of expression (such as [4]), we specifically chose to focus our tool on sensors as the means of input, and lights as the primary means of output for garments. This is because in the 2.5 years of our work with MakeFashion, lights and sensors were the most common form of input and output chosen by design teams for their garments shown on their runways, due to their visual nature. Additionally, teams working with illuminated materials (and their interactions) contributed to some of the challenges we observed. Mannequette also facilitates communication and creativity between the team members through a shared and tangible medium of communication. Next, we describe our design principles for tools within the avant-garde fashiontech space and our Mannequette system, and how it works.

Design Principles and Goals
Informed by prior informal interviews through our relationship with MakeFashion, as well as our experiences with avant-garde fashion-tech runways, and their associated technologies (including [1,5,6,39]), as well as relevant prior work (e.g., [17,29,34,37]), we synthesized the following key goals for a tool designed to facilitate communication and creativity between multi-disciplinary teams: Leverage existing models and processes. While several toolkits for wearables exist and provide a rich number of components [2,39,42], many don't consider the creative and construction processes within avant-garde (for both designers and technologists) or runway environments. In contrast, we aim to focus our tool on leveraging specific processes that exist within avant-garde and fashion -for example, a tool could be designed around a mannequin (and the practice of maquette, which plays a role in the areas of ideation, prototyping and discussion in art and fashion. Tinkerability. Avant-garde fashion-tech teams rely on their combined team skills, supplemented with additional skills from others (e.g. craft, fabrication, design, programming) to create a garment. Similar to wearable toolkits for children, emphasis should be placed on allowing quick tinkering and prototyping regardless of available skills on a team [34,37], especially in the areas of lighting (LED) patterns and sensors. Tinkering occurs heavily in the early phases of the garment construction process we observed (e.g. when choosing fabrics), so tools must be able to accommodate for a wide range of tinkering tasks within the construction process.
Low floors, high ceilings, wide walls. Building upon design cues from [17,29], tools must support teams in the creation of increasingly complex designs as they gain experience. Not all avant-garde fashion-tech teams have technical experience in relation to garment construction, so accommodating differing levels of skills with multi-disciplinary teams is important in designing a tool.
Augmenting avant-garde. Avant-garde fashion-tech is heavily focused on aesthetics, story and culture, much more than traditional wearables which focus on practicality. Given the highly experimental nature of Avant-garde, we aim to support and augment as wide a variety of designs as possible that emphasize look (and occasionally simplicity) while also allowing for creative exploration and incorporation of sensors and lighting in a meaningful way.
Modularity. Modular wearable kits and systems have proven to be successful in the past [17]. For the processes we observed, where dress forms, fabrics, and tools are continually swapped, building upon these principles can provide value. Fashion designers (and technologists) should be able to quickly swap between components (i.e. sensors and lights) and interactions. Additionally, using a modular approach from a fashion and technical perspective, allows a team to collaboratively create garments in a more systematic manner (e.g. a dedicated and removeable technology layer, removeable casing for wiring). This also helps with adjustment or repairs for a garment if and when technology breaks down in the garment (e.g. loose wiring, or an electrical short), compared to unsewing or entirely redoing parts of a garment. Furthermore, for runway conditions when the environment is stressful for a design team, a modular system can enable quicker and easier debugging.

Mannequette System
Based upon our design goals, we created Mannequette (Figure 3a), which is comprised of three parts: (iii) a custom I²C-based DJ mixer-like interface that allows designers and technologists to dynamically prototype behaviors with (virtual) sensors and light patterns, which can then be integrated and used by a proper physical sensor in the final garment; (ii) a tech-infused base that supports a plug and play system of grove-based I²C sensors which are used to create interactive behaviors; (iii) a swappable miniaturized dress forms. The dress forms are 3D printed and scaled from open source models, and minimal modification is required for additional 3D printed dress form models.
Mixer. The mixer control panel for Mannequette serves as the hub for teams to communicate and quickly experiment with a number of light patterns and interactions. Inside is a Seeeduino Lotus, 4 Seeed Grove I²C-based buttons and 2 Seeed Grove I²C-based potentiometers, enclosed in a custom-designed 3D-printed case (Figure 3b). Each button and slider is mapped to specific functionality regarding a pattern and interaction: saving, recording, play back, and loads/saves/clear, as well as adjusting brightness and sensor values ( Figure 4). The control panel connects to the base using an I²C cable, and a custom I²C communication protocol that was written to enable customization of patterns and interactions that are saved/loaded/cleared onto the tech base. We initially explored using a mobile application, as well as a WIFI or Bluetooth-based approaches (e.g. a mobile phone app) but we aimed to support the full set of activities described earlier, particularly runway environments where debugging is critical and Bluetooth and WIFI does not work sufficiently -due to changing or noisy environments-based on our experiences.

Tech Base and Dress Forms.
The tech base consists of a customized ATMega32U4-based Arduino board inside a custom 3D printed casing, with a slot on its top where the dress form holder snaps into place, with swappable dress forms. The tech base supports 2 outputs for WS2812b (or similar) LED strips and 1 I²C port, which the mixer uses to program the tech base and is also used for the sensors. The information coded by the mixer (e.g. light patterns and   Supported Sensors. Selecting and abstracting a set of sensors is crucial for enabling teams to create garments. In addition to supporting standard modules used by makers who create wearables (e.g., WS2812b LED strips for lighting), we also focused on supporting sensors that could quickly assist in interactivity for the runway environment [44]. A runway environment constrains the effectiveness of certain sensors (e.g. Light or Sound) and communication technologies (e.g. Bluetooth or WIFI) for garments because as they travel, these conditions cannot be guaranteed, nor is it guaranteed that the team handling a garment at a show will be the same as the original design team. Thus, we emphasized sensors and interactions we knew have worked previously in runway conditions, by examining 5 years' worth of garments from MakeFashion teams, as well as informal interviews. We supported: body movement (proximity, motion, accelerometer), environment (light, sound, temperature, heart) and input (touch, button, potentiometer and rotation) to start. Our intention was not to immediately explore all sensor possibilities, but to instead begin with those frequently used or requested by teams in MakeFashion.

Prototyping and Finalizing Interactions.
Creating different sensor and light interactions generally occurs by using the mixer (see Figure 5). First, a team chooses a predefined light pattern from the mixer. We provided 23 predefined (but customizable) light patterns chosen based on our prior examination of garments created in MakeFashion. The brightness of light patterns can be adjusted by using the brightness slider on the mixer. After the team (or specifically the fashion designer) has completed experimenting with different light patterns, they configure different sensor interactions. Sensor interactions are mapped to the sensor slider where the values of sensors are mapped to a value in a range. For example, a button (or touch) interaction is mapped to 'off' (or 0) on the slider, while 'on' (or 100) is mapped to maximum on the slider. Similarly, for sensors such as light, sound or proximity, ranges can be created for low, medium, and high. For each value (or range) on the slider, an associated light pattern can be recorded and played back. If the team does not like what has been recorded, it can be cleared, or they can save their creation for further (or final) modifications on a garment. The mixer supports 5 ranges.
After a design team is satisfied with their prototyped interaction and it has been saved, the light patterns can be further customized, and a supported sensor can be used, as well as tweaked. To modify a light pattern, the tech base is plugged into a computer running Arduino, and a pre-defined region of Arduino code for its color and other properties (e.g. speed, delay, number of pixels) is modified. Each of the 23 light patterns we provide are modifiable with their own properties. Additionally, the light patterns are templated in a manner such that additional light patterns can be added to the Mannequette system by a technologist (or designer) if they chose. To use a sensor with the interaction that was recorded and saved, the associated sensor is plugged into the I²C port of the tech base. As the sensor values were mapped to the interaction slider, if the sensor values need to be tweaked (e.g. volume level or distance), a predefined section of provided Arduino code can be adjusted by a technologist (or fashion designer) to tweak these recorded values.

DEPLOYMENT STUDY
To gain insight into how multi-disciplinary teams consisting of (self-identified) fashion designers and technologists can use, understand and communicate with Mannequette throughout the process of creating an avant-garde fashiontech garment, we conducted an 8-week deployment study.

Method
We recruited eight teams through our partnership with MakeFashion, which hosted an avant-garde fashion-tech runaway show in Fall 2018. Each team was self-selected and varied in experience with avant-garde fashion-tech (both garment and technical), as well as the number of (selfidentified) technologists/designers for the team. We also asked each team their perceived level of expertise with technology in fashion, with respect to areas such as 3D printing, laser cutting, electronics and programming. They were asked to rank this in terms of limited, medium and strong experience. Table 1 describes these teams in detail.

Procedure
Over the course of the 8-week period, we engaged our design teams in bi-weekly interviews and discussions, tracking their progress and processes of constructing their garments. We also asked teams to document their creation process. When teams completed their garments and felt they were runway ready (i.e. ready to be worn, technology had been integrated/tweaked and ready for dress rehearsal), we ran a final interview. We did not track team progress beyond the first appearance of garments on the runway.
We deployed a Mannequette to each of these teams following their selection to feature in their annual runway show. We provided each team with the Mannequette system, a 5cm strip of WS2812b LEDs and one USB-C cable for the tech-base. Teams were given 10 tutorial videos for Mannequette that covered topics including assembly, how to use the mixer and how to modify pre-defined code for sensors and lights. Finally, teams were provided sensors and additional LED lighting strips by MakeFashion when needed for experimentation, as well as for their final garments.

Data and Analysis
We employed a mixed-methods approach to assess the effectiveness of Mannequette, as well as the challenges teams faced when constructing their garments using Mannequette. The first interview with design teams collected information about prior experience in avant-garde fashiontech, and the last collected their overall experience, and how well they felt they accomplished their design (and technical) goals with Mannequette. All interviews were captured and transcribed, and we used a thematic analysis approach to analyze our interview data.

Findings
In total, the eight design teams created a total of 15 garments using Mannequette. Table 2 also provides a brief breakdown of the garments each team constructed and the types of interactions (i.e. sensors) used in the garments.
We first frame our findings in the context of the process described earlier (and MakeFashion) and discuss themes and common patterns we observed with teams' use of the to communicate ideas about more complex temporal patterns to technologists. These could be visualized with the Mannequette and articulated in-place with the tool, rather than being described verbally or through paper sketches. Some teams even shared their patterns with other teams who were not using Mannequette but were also creating garments for the MakeFashion show. Thus, these videos of the Mannequette acted as inspiration for other teams. A Team 3 member: "…I thought the pattern I made was very pretty and something that [another] team could modify, since they make interesting things and it would make theirs so much cooler." We also observed teams draping fabrics and components of using the dress form of the Mannequette. This allowed them to see how the fabrics would interact with sensors and LEDs early in the design process, as well as plan for the placement of components and batteries. Unlike a traditional mannequin, however, the Mannequette did not allow teams to pin fabrics. A designer from Team 3 stated: "…the miniature size of the dress forms is great, but we can't pin to it directly which makes it a bit challenging sometimes." Experimenting with Mannequette. Prior to the introduction of Mannequette, MakeFashion teams only began experimenting with technology late in the process (i.e. distinctively after working through the inspiration phase). With Mannequette, this process changed dramatically: teams (especially fashion designers) experimented with different interaction concepts very early on in their design process, where the delineation between the inspiration and experimentation phases was considerably less distinct. This seems to have been made possible by how Mannequette makes interacting and testing ideas easier.
For example, we observed teams working collaboratively to rapidly iterate through sensor interactions on the miniature dress form ( Figure 6 The Runway with Mannequette. Teams used the Mannequette (specifically the mixer) to fine-tune their garments for runway conditions, and those with more complex interactions and sensors benefited the most from having the Mannequette. For example, Team 6 (consisting of a single person who was both a designer and technologist) used a proximity sensor, where the concept was that anytime someone approached the model, their garment would react. The proxemic interactions were prototyped and constructed without a clear sense of the runway size, timing for music, and distance that the garment would need to react with another model on the runway. During the dress rehearsal, where garments were tested, this team adjusted their proxemic values ad-hoc using the Mannequette while on the    We also observed teams begin to introduce modular concepts into their design practices and how they fixed garments. For example, Team 3, who specifically designed a removeable technology layer, found a small short in the wiring for an LED strip. This was discovered after a model wore the garment. As they describe: "…we're really glad we caught this at the rehearsal and that we could diagnose it so quickly and fix it the way we did. Having the mixer allowed us to isolate the problem with the help of a technologist."

DISCUSSION
Mannequette provides one solution for some of the issues we described in the avant-garde fashion-tech process, but it is not a silver bullet. Our aim was to introduce a complementary tool within this process to aid in overcoming some key challenges in designing the garmentscommunication and creativity. We discuss the key takeaways when designing tools like Mannequette within the broader scope of wearables, as well as the impact of facilitated communication within avant-garde fashion-tech.
Redesigning Tools for Wearables. Future tools for designing wearables must be inclusive to other design processes and communities to enable richer explorations [38]. While several toolkits for wearable fashion exist, they seem overly focused on maker communities or education. And, although these toolkits have been wildly popular within those communities, they have had comparatively limited uptake in the avant-garde fashion-tech community, particularly because they do not consider pragmatic concerns such as weight, robustness, and power. Tools in the fashion-tech space can incorporate lessons learned from prior work in construction kits [4,5,18,25] to lower the barriers of entry into technical areas. With Mannequette, we still observed some discomfort in programming by designers, despite some becoming interested or even empowered in those activities.

Within and Beyond the Design Team.
Maker communities often openly share designs and collaborate, with novices able to build from designs depending on what tools are available to them [12]. With Mannequette, we observed a small step towards this within avant-garde culture, as many designers became increasingly confident and empowered in communicating their ideas with technologists. Designers began to use their own documentation as a means of explaining and comparing temporal, and also demonstrating their work to others (through video and pictures). Several even began to share their creations and construction processes to other designs teams in MakeFashion, as well as the broader avant-garde fashion-tech community through social media (e.g. Twitter, Instagram). This is important because one of the main challenges in the observed process discussed earlier was the lack of examples for novice teams to draw upon. Furthermore, the video and photo documentation created and shared can be used as a means of documentation for the garment itself, especially useful when the garment is handled by other designers and technologists at different runway shows. Ultimately, we envision this approach of using a tool to create expressive visual artifacts, coupled with documentation processes, a valuable way to grow the avant-garde fashion-tech community, similar to [5].

LIMITATIONS AND FUTURE WORK
We discuss the limitations of our work and suggest future research directions for the avant-garde fashion-tech space.
Design Modularity -Our concept primarily focused on creativity and enabling interactions, but we did not fully explore the modular aspects of our design and its benefits in extensive detail. Furthermore, we used existing off the shelf components ( [44]) which were already modular but not always ideal for garments (e.g. bulky connectors).
Team Dynamics -The lead author was deeply embedded in the MakeFashion organization, which has unique team dynamics between (self-identified) fashion designers and technologists. MakeFashion specifically creates or finds teams that contain fashion designers and technologists who create garments. Much of our observed and described processes is built upon this, and we recognize this may not be reflected in other organizations and runways. In our future work, we will explore working with other organizations that may have different team dynamics and processes.
Evaluation -While we deployed Mannequette in a realworld setting with design teams, our period of evaluation was limited. We followed the progress of teams from concept to execution, but we were not able to explore the later phases when the garment travels along with its Mannequette.
Alternative Forms-While Mannequette focused on using a whole-body mannequin for prototyping, other singular forms do exist, such as arms, or heads. Exploring these forms individual may enable designers to have even more interesting experimentations and explorations into wearable technologies. We also focused on normative binary body shapes for this initial work, but our future work will explore using different shapes of different bodies for designers.

CONCLUSION
We introduce a prototyping tool for avant-garde fashion-tech garments, designed to address the challenges that arise from designing and experimenting with interactions, sensors and outputs at the earliest stages of the design process. It does this by providing different detachable dress forms, support for different sensors, and a custom baseboard paired with a simple DJ mixer. We first embedded ourselves within a fashion-tech organization, before designing and building our tool, Mannequette. We conducted an 8-week deployment study with eight teams who designed garments, from concept to runway. Our results provide insight into how to design tools that incorporate the avant-garde process and facilitate interdisciplinary communication and creativity.