Let's Play Something New!: Designing for Digital Malleability in Outdoor Playgrounds

Outdoor playground design has traditionally involved building installations that after inauguration remain stationary and rigid. New technologies open possibilities to reinvent playground design, but also provide new challenges. This paper focuses on how an outdoor playground can be made malleable through digital technology, and how such a design can increase play values and provide more flexible possibilities for an iterative design process, also after inauguration. During a total period of 5 years, using a Research-through-Design, and more specifically Reflective Design, approach researchers participated in a co-design project where a hybrid Internet of Things-enhanced permanent outdoor playground was developed and studied. The installations have been up and running around the clock, and used daily, for 2.5 years. The study suggests ways to design digitally malleable playground installations in permanent playgrounds. Further it points to areas where malleability functions may be restricted, due to reasons such as municipality policy, security and resources. It shows that the malleability functions were mainly used to increase usability, for educational purposes, to enhance play values, and to allow for appropriation and co-creation. The malleability features enabled an interactive design after design approach, and those possibilities were utilised in several ways, adding values to the playground, compared to rigid playground designs.


INTRODUCTION
Outdoor playgrounds are a global phenomenon, and playgrounds look surprisingly similar in most places [31,47].Industry playground design is often conservative, and tends to lack innovation [80].It is common to build moulded in place rigid installations, so called Kit, Fence, Carpet-playgrounds, even though they do not cater well for variation of play [19,50,80].These types of rigid playground designs are common, despite that loose materials and malleability are well known to provide important play values and play possibilities [54,56].According to several leading play scholars such as Sicart, Piaget and Sobel, to be able to change and appropriate is a core element of play [59,69,70].Therefore, it is important to design playgrounds that provide ample possibilities for appropriation and malleability.
Recently, digitally enhanced playground installations have started to emerge [38], although this is still a young phenomenon.Digitally enhanced installations offered from the industry are often very simple from a technology perspective; buttons, LEDs and mini games [51,60].Often, what the digital layer can do in those installations is pre-determined by the developer.Further, they are often rigid in their experience design.Even though a digital layer is added, they do the same thing all the time.However, connectivity and the Internet of Things (IoT) can offer interesting possibilities to make play installations more malleable, as is pointed out by the industry magazine Playground professionals, where IoT has been addressed as one of the emerging technologies that will affect playground design the most [38].In this work we explore how public outdoor playgrounds can be designed to be malleable, using digital layers in physical installations, and how those malleability possibilities can be utilised, and what values and challenges this approach comes with it.We define malleability as the capability of an artifact to be influenced and altered, in order to change affordances, functionality and/or the possible interactions provided by the artifact.We deliberately stay clear of separating between digital and physical in the definition, to open up the design space for working in hybrid formats where both product design and programming are important.Further, we do not make an absolute separation between designers and users, this way we open up for several perspectives on the design work.
We take on a Research-through-Design approach, as formulated by Zimmerman et al [82], designing digitally malleable playground installations to apply malleability into a traditional (rigid) playground design, in order to increase the intrinsic values provided by the playground, and investigate possibilities provided by this approach.Playground designs involve many practical considerations, such as safety aspects, regulations and laws, adapting to public concerns and so on.These all affect the design choices [80].Therefore, prototype designs can differ a lot from what would be possible when designing for in-the-wild use and implementation.Here, we work with the design of a permanent playground, meaning having to adapt design choices to the practical and policy realities of playground design.The research has been conducted by designing and studying several testbed play installations over a period of 5 years (2.5 years of designing the installations and 2.5 years of studies since inauguration of the playground).By working 'in-the-wild' we have been able to take into consideration not only design challenges and implications that comes with prototyping, but the ecosystem of requirements for an authentic case.
The playground has been informally and formally evaluated in multiple studies.In this article, focus lies on how malleability supports a 'design after design' process, and the changes made to the installations over time from when they were first installed.Our main source of data is for that reason not the children's experience of the playground, but the changes that were made to the digital content over time, by both children and other stakeholders.From this we elicit some key features that support malleability in interactive playground design.

BACKGROUND
During the early 20th century the modern playground evolved as a separate outdoor space filled with play installations such as swings, slides and sandboxes.The foundation for generic playground design developed [21].Within play research, both the intrinsic values of play, as well as pedagogical, physical and psychological benefits have long been known [59,78].Still, most contemporary playgrounds look surprisingly similar to those early 20th century playgrounds, and playground design tends to be conservative in their design for several reasons; including policies, economical tradeoffs, push for standardisation, safety regulations, laws, parental concerns, and fear of accidents [80].Today playground equipment is a large global industry, its market size being USD 5170.1 million in 2021 forecasted to grow to USD 8275.1 million in 2028 [45].The playground industry is dominated by a few large global companies, many children therefore play in mass produced playgrounds that lack connection to the local environment [31] and that tend to look quite similar, designed without involving the children supposed to play there [47] and without the needs of the children in focus [81].Roger Hart, professor in psychology with focus on children's play, argues that modern rigid playgrounds fail to cater for children's need for play [30], and only encourage certain types of play, such as physical play.

Design Practises in Traditional Playgrounds
Playgrounds are big building projects with long planning time and life expectancy.They must be designed to endure tough weather and children playing all year round.Typically they are built to last ten years or more [62], and to stand without the need for daily supervision and with a minimum of operating and maintenance costs [35].Constructing a playground poses specific challenges that affect the design choices, typically involving laws and regulations concerning public playgrounds [35,52], the need for a very high level of robustness due to wear and tear from the children, weather, and the fact that the playgrounds for most parts are unsupervised.All those challenges must be taken into consideration also when designing digitally enhanced playground installations.Traditional playground design tends to favour physical play, and missing some advantages of more creative playgrounds, and playing in natural environments [44,76].

Digitally Enhanced Playgrounds
While some digitally enhanced playgrounds have evolved during the last years, digital enhancement of playgrounds are still in its early stages [20].Currently digital enhancement is usually in the form of non-connected installations with all technology inbuilt, often with LED-lights and robust buttons [60], or as playground apps [61] (often in combination with QR-codes) intended to be used together with your phone.There are a few examples of technology enhanced playground installations offering choice, such as the Sutu [40], where the children can choose between different ball games, and the FONO DJ-table [41], that allows users to play their own music using their mobile phone.And there are examples of connectedness.For example, in mars 2023, Virgin Media opened a temporary connected playground in Islington, describing it as "The world's first connected playground" and a "glimpse into the future of children's play" [14].The playground was broadband connected with interactive installations using sound, light and sensors, but to use it you had to book time slots, and the playground was supervised for the two days during which it was open for the public [13,14].
There are examples of digitally enhanced playground catalogue products [51,60].Those examples tend to have simple technology solutions, with robustness as an important design quality in focus.The rules and regulations for playgrounds are not adapted for digital enhancements, and neither are the traditional design process.Therefore, clients need to adapt the design and building process, for instance regarding adding electricity cables, internet connection and so on.Also, the operation and maintenance budgets for playgrounds is often low, only for repairing physical damage, such as a broken swing.Digital technology provides a whole new layer of possibilities but also challenges.Those challenges increase even more if the digital technology is supposed to be adaptable.
Despite these obstacles, the industry magazine Playground Professionals [38], foresees that the Internet of Things will become a trend in playground design, and sees that IoT will hugely change how playgrounds are designed.According to the magazine, benefits could for instance be enabling location-based playground games, and increased security.This paper takes on another approach, seeing how Internet of Playground Things can provide play values by offering malleability, and enabling appropriation through digital layers.

Playgrounds, Play Values, and Value of Play
Playfulness can be seen as a core human trait, something that has been acknowledged by many [1,11,12,34,68,69].Play is widely acknowledged to be a central part of childhood, and to the utmost importance of child development, with multiple classical theories and studies on the benefits of play, ranging from physical development, learning social skills, emotional and intellectual development, recreational purposes and much more [4,21,59,78].Play is also acknowledged by the UN as a fundamental right, in the Convention of the right of the child, article 31 [83], and public outdoor play, for instance at playgrounds, is an important form of play [31,31,47,50,56,80].But children's access to free public play, especially in the western world, is declining [36].Children, at least in the western world, today are very restricted in where they play in public environments.According to McKedrick there is today a hidden contract, that children should play at the playgrounds and not elsewhere in the public space [47].Critique is lifted against the common designs of public playgrounds today, for not corresponding to children's need for outdoor play [36].Also, we can see a commercialization of play and childhood, but without giving consumer power to the children, rather they become a commodity [48].We can see play deprivation for children, especially for unstructured outdoor play [4,43].At the same time there is a move towards acknowledging the importance of play, not only for children and in designated places, but also for everyone and everywhere, such as the concept of the playable city [1,55].There is a push in design to move away from the efficiency paradigm to allow for more playfulness [1,27], for instance by moving away from gamification and its focus on achievements [3], to instead focus on playful exploration and playification [5,46].Designing for public play and playfulness can be of importance for several reasons [1,2,9,25,39,49,55,57,58,65,71,74,75], for instance in creating a stronger community [72], increased civic engagement [25] and creating social flow [15].
There are many identified important factors to create attractive play environments, including for instance the ability to change them, constant development, children's possibilities to control and manipulate the space, user involvement and site specific adaption [36], all relevant in relation to cater for malleability at the playground.Hallmarks of a good play environment are not obvious, but one way of approaching it is via the concept of play values [29].There is not just one set definition or list for play values, rather there are several related approaches and taxonomies to identify toys, play installations and situations that provide interesting play opportunities, especially for children.Hughes identified a taxonomy of play types [33], including symbolic play, locomotor play, mastery play, exploratory play, rough and tumble play, social play, recapitulative play, creative play, fantasy play, role play, socio-dramatic play, communication play, dramatic play, imaginative play, object play and deep play [33].Inspired by Hughes, among others, and important in the field of playground design research has been the TOPO, (tool for observing outdoor play) that provided a typology for outdoor play [42].Yet another way of looking at play values is what Bertran et al describes as play potentials, among those especially relevant in relation to the malleable playground are Hack the street, Asynchronous communication, Urban challenges and Temporal decontextualization [1].Common among all is an understanding of play as a broad phenomenon, treating play as something that can take many forms.Meaning, a play space that offers a wide variety of play types, as well as good opportunities to indulge in each type of play, could be considered to carry more play values.

Loose Materials, and Playful Appropriation
Play values are not only about offering different play types.It is also for instance about the possibility to find new play opportunities, and for the children to be able to find and create such opportunities.A playground that offers loose materials such as sand, water and pinecones, nature play, and a varied landscape that will change with seasons and over the day, creates such possibilities.Such loose materials are known to be important for play values [54,56], as they give children the opportunity to be co-creators of the space, as well as letting the space itself change and evolve.Different types of explorative and discovery play presuppose that there is some change happening, for a child to be able to discover new things also at a playground that they visit regularly.
According to play researcher Sicart, appropriation is a key element of play [69].The influential psychologist Piaget, focusing on child development, stated that play is about appropriation of objects, playfully appropriating them and thus through play experimenting to understand how the world works, and creating a foundation for emotional stability [24].According to Sobel, appropriation of play spaces, making them one's own, creates a sense of agency and ownership [70].Playful appropriation is also important for negotiating space, and becomes a way to experiment with conflict management and power structures.Such negotiations of power through play and constructive conflicts can be beneficial for emotional and social development [16].To be able to appropriate is about making something one's own.For this it is important to be able to affect the space.
Seeing appropriation as a key component of meaningful play interactions, and then contrasting this with the widespread rigid Kit-Fence-Carpet playground designs, points to a need for change in current approach to playground design.According to the United nations convention on the rights of the child, children should have the right to express their view on matters that concern them [83].And what is playground design, if not something that concerns the children?A large interview study by Jansson suggests that implementing children's perspective into playground management and design can improve play possibilities.The study especially calls for the importance of placemaking and manipulations as important values at playgrounds [37].

Time, Children's Involvement, and Appropriation
A challenge with involving children in playground design is that building playgrounds takes time.Even when children are involved, those children will have grown past their play period before the playground is finished.This points to the need for fast adaptations to give the chance for children to manipulate and affect their surroundings also with an already built playground.While designing together with children can be fruitful and give new ideas, it can also be complicated as ideas often are not grounded in practical reality, and group dynamics may be troublesome.This has been pointed out before by researchers working on co-design with children, e.g. in interviews with designers [73] and documentation of design work [77].One approach is to prime for malleability and appropriation already in the design process.Even though appropriation largely is about breaking the intent of design, it is possible to deliberately design for appropriation.For example Alan Dix [17] presents a guideline framework for how to design for appropriation, consisting of 7 principles, namely: Allow interpretation, provide visibility, expose intentions, support not control, pluggability and configuration, encourage sharing, learn from appropriation, and Back et.al.[7] points to deliberately treating the user as a co-designer, giving them the freedom to appropriate through re-design in use.
Allowing interpretation and different uses also relates to designing for ambiguity, as described by Gaver et al [23], seeing it as a value, not a drawback, ambiguity becomes extra interesting in relation to having the same installations offering different play values and opportunities.In this paper we focus on how a digital layer can provide possibilities for malleability and appropriation.With increased entanglement of the digital and physical, we need to rethink the relation between the digital and physical public space when designing for social interaction [18].Spaces can be both digital and physical, and turned into places by being filled with cultural meaning and by appropriation [28].This relates to the concept of creating the playable city [55], to allow for uses of technology, such as in this case IoT, to enable opportunities to hack spaces and make room for citizens (also the youngest ones), to repurpose spaces, appropriating it through technology.But this paper is more literal, not focusing on the city as a digital playground, but rather focusing on the digital in actual playgrounds.

METHOD
This paper builds on a research-through-design (RtD) approach as formulated by Zimmerman et al [82], The RtD approach used is grounded in Reflective Design, as formulated by Sengers et al [67].
The viewpoint is that RtD are, and should be, emergent, contextual and transitional [22,63].We as researchers participated as full design team members in a design process, designing a playground in a social innovation project.The project, as well as the research, took on an In-the-Wild [8,10,64] approach, studying the "naturally occurring, culturally constituted human activity" [8] while designing an actual permanent playground, not just prototypes.Similarly, observations and evaluations were done on actual user cases, not in an experimental setting.It may be pointed out that when we say in-the-wild, we do not necessarily mean the wilderness.The authors followed the project for three years, and took active part in the design process.This paper aims at producing intermediate level knowledge, as formulated by Höök and Löwgren [32].While the designs themselves may be treated as ultimate particulars [53], we are emphasising the particulars not mainly as artifacts, but rather as how humans interact with each other through the designs [79].

Data Gathering and Analysis
During the design process we as researchers participated in most design meetings, we also gathered data from the design process in the form of blueprints, photos, design documents and presentations.Workshops, prototype testing and user studies were specifically documented with photos, notes and in some cases video.Further, the developed designs as well as the prototypes are in themselves considered bearers of knowledge.The finished physical play installations, as well as hardware and software, have been documented with pictures, videos, blueprints, measurements and technical specifications.Also, new games and scenarios developed have been saved, and tweaks and changes that have been made after inauguration have been documented.After inauguration several in-the-wild studies [10,64] of play behaviour have been conducted, this was done through a combination of observations and semi-structured interviews.No personal data was collected during the observations, and for the interviews consent forms were used, as well as discussing the consent with both children and caretakers.Formal observations at the playground were conducted in September 2020, December 2020, January 2021, February 2021 and May 2021.For each study at least one observer, but most often two or more, observed the playground during a couple of hours each time, focussing on documenting detailed description of play behaviour.In some cases, the children and their caretakers were approached and informed about the project, and if they gave consent a short interview was conducted on site and documented through note-taking.During the interview the children were asked to tell and show where they had just played, what they had been playing (describing the play/game) and how it had felt.Apart from the formal observation sessions visits have also been done to the playground for different reasons, often several times a month, since the inauguration.For instance, in deployment of new scenarios and features, when helping students with their design projects, and to check when malfunctions have been reported.Further, one of the authors partook in all the code summer camps, and anonymous feedback was also gathered by the summer camp organisers, and shared with the project team.
Changes made to the installations after inauguration were monitored by the researchers for two and a half years.After excluding changes due to bug fixing, correcting malfunctions, tweaking to already implemented scenarios, and temporary implementations such as tests of functionality, 11 cases of malleability were identified.The malleability examples were coded and clustered based on who were involved in the design (designers, project members, children, students) and based on the reasons given by the design team for using the malleability functionality (usability, play possibilities, children involvement, learning situations), they were also related to what play values they intended to increase.Key examples were chosen and studied in detail and design challenges and possibilities were identified for the chosen examples.These are presented in the results.

About the Design Project
The design project in this paper aimed at creating Internet of Thingsenhanced playground installations that were developed as part of a permanent outdoor playground.The playground installations discussed were developed in a state funded innovation project, focusing on Internet of Things and playfulness, the project ran for 3 years (2018-2021).The project consortium included a municipality, a tech company, an educational company focusing on playful learning for sustainable development and two universities, in both cases engaging HCI researchers.The core of the design team referred to in this paper consisted of personnel from the tech company, both hardware and software experts but without previous experience of playground design (one of them an author of this paper), a municipality landscape architect specialized on playgrounds, a municipality representative responsible for the city's gardens and outdoor spaces, a designer and a project leader from the educational company with a focus on playful learning experiences and sustainable development (also, an author of this paper), and a design researcher from one of the universities with a focus on playful outdoor design work (author).This core group met on a weekly basis, and participated in design workshops and other relevant meetings.Apart from this core group there was also an extended group involved in the design, including for instance personnel from municipality, builders, other researchers, invited ideation experts, an author and an editor.The project was heavily value driven, focusing on a range of values to integrate into the design, working with six identified core values in accordance with [6].In this article, we focus on the core value 'malleability for appropriation' (adapted from the referenced article's 'changeability'), and design aspects and values relating to that.

THE INSTALLATIONS -DESCRIPTION AND IMPLEMENTATION
The installations that are described below are located in a public outdoor playground in Linköping municipality, available for all citizens to access and play with.The interactive playground (see Figure 1) is one part of the playground which also consists of swings, climbing installations, a padel court, a barbecue area, and a small soccer field.The interactive playground is situated at a corner of the playground, close to a small forest with the intention to blend in with nature.The playground is permanent, so the installations had to be built to be 1) Secure: both when it comes to physical security as well as the IT perspective.2) Robust: need to withstand the weather of all seasons (from +40C to -25C), handle internet outages, the physical interaction of children, etc. 3) Accessible: for citizens who are limited by disabilities.The playground was inaugurated on the 19th of September 2020.The playground technology and the installations have been running around the clock for 2.5 years (at the time of writing).There has been some maintenance needed, but only to a very small extent.The gravel around the Eggs have had to be refilled several times.The worn-down gravel happened (according to the maintenance staff) because of the huge number of children running around the installations.During a logging period of 2 weeks over 2000 play scenarios were activated.
There are two types of installations in the playground, the eggs and the hut (see Figure 2): The eggs (two of them) are approximately one metre tall and designed to mimic a natural wild bird egg.Inside the eggs there are two marine loudspeakers, LED lights grouped into 16 light points.16 capacitive proximity sensors that can detect a human body at short range without responding to small objects or dry organic matter, one at each LED point.
The hut has a similar configuration of devices as the eggs.The hut consists of four tree stump stools, which contain one capacitive sensor each.Above each stump there is a LED spotlight mounted in the ceiling, with the light cone directed at the stool.There are four speakers mounted within the inner walls of the hut, resulting in a stereo sound with two loudspeakers per channel.Finally, there is a LED strip in the ridge of the ceiling, with eight separate sections which can be operated individually.All electrical devices are connected and mounted inside the installations and controllable through software interfaces over the TCP/IP protocol.

Software and Servers
The system consists of two subsystems, the playground server (located in an electrical cabinet in the playground) and the playground service (located in the cloud).The two subsystems communicate via an internet connection.The entire system was designed with the idea that interactions should be adaptable through a block programming interface.The framework Blockly [26] was chosen as the foundation for the programming interface, which is also used in other contexts such as the popular tool Scratch [66], made for introducing beginners to programming.As part of the playground service subsystem a web application called "Scenario Designer" was constructed to enable the programming of the interactions in and between the eggs and the hut (see Figure 2).In the Scenario Designer the user is given access to the devices in the playground installations and can use them for designing their own interactions.The scenarios designed in the Scenario Designer are saved to a database and are also downloadable as XML files.When a scenario is deployed to the playground it is transferred to the network connected playground server which is located in an electrical cabinet close to the playground installations.The system uses the scenario code to generate javascript code, which is then executed by the server when a scenario is run.Once uploaded the scenarios will keep running as long as the playground has power, even if the internet connection is lost.
4.1.1Interaction Design for the Installations.Several scenarios were developed prior to inauguration.In the eggs, there was a startup/paus scenario from which the other scenarios could be chosen.Each scenario had a colour, and if you touched a certain coloured sensor, the scenario of that colour would be activated.This meant that new scenarios could be added to the eggs without changing the old ones, and that specific scenarios could be removed without interfering with others.Each scenario is a short game-like interaction with its own rules and interaction pattern.Scenarios were designed to encourage collaboration and offer a variety of play types.At inauguration the scenarios were: 1) "Paint the egg" -By touching a sensor, that location changed colour.2) "Hunt the dot" -A dot lit up, when touching it, the dot moved to another location, and so on.3) "Keep alive" -The egg light up, then colours fade.As you touch the sensors, they light up again.
At this point there was only one scenario for the hut, and therefore no specific startup scenario for selecting was needed.This scenario was named "story mode" and consisted of listening to stories about the different animals living in the forest.The stories are written by a children's book author and narrated by voice actors, 295 story snippets in total, each around 1 min long.The stories were written in 4 sections of narrative: summer, fall, winter and spring.Which narrative was available in the hut changed with the actual season.There were 8 different animals followed in the stories (each narrated by a different voice actor).By changing seats, listeners could choose which animals' perspective to listen to.Each animal was colour coded so that coloured spotlights and sound indicated when to change stool.

MALLEABILITY OF THE PLAYGROUND INSTALLATIONS
In the data, three perspectives were identified: types of malleability, who made the change, and types of play-values added.The perspectives will be shortly described below.As all perspectives overlap, and as the project has been ongoing for several years, the categories are not always clear, and the knowledge is combined with gained understanding from throughout the project.Therefore, after introducing the perspectives, five key examples will be used to show more detail, and reference back to the perspectives.

Types of Malleability
When studying in what way the malleability was explicated, two main themes were found: everyday malleability, and active reprogramming.These will be described below."They have started the blue scenario and many lights are lit already.They touch the lights and the younger sister says "it should be pink", many of the lights are already pink, but grandma accidentally changes one of the pink lights to another colour "oh sorry" says grandma and looks at grandpa and they laugh.The children are happy and sometimes laugh and say little things to each other.The big sister says she too thinks the pink ones are nice and they push to change them all to the same colour." (Translated observation notes from field observations January 2021) The following quote shows both the confusion caused by the malleability, but also that it is appreciated.Especially note that the interviewed father stresses that the playground now is different since last visit: "Later, the father comes over and asks how the games at the eggs work.They haven't been here since the "new games" came and he wonders how they work and how to choose."We didn't really understand"."I think it's cool with this interactivity in outdoor play, [name of the child] especially likes the fairy tales very much" "We've never seen anything like it before"."(Translated observation notes from field observations January 2021) 5.1.2Active Reprogramming.Apart from the ongoing above examples of malleability, during the two years after inauguration the playground was changed several times through active reprogramming.As an example, the four different software components in the system have been released 15 different times after the initial version 1.0 of each component.This shows that the possibilities provided by the malleability functions were used.Changes were made by different stakeholders and for different purposes.Also, after the project was finished (and the playground became part of ongoing operation of the municipality, now lacking the project funding) the malleability functions continued to be used, most notably by yearly Code Summer Camps.There are also plans (in June 2023) from the municipality to use the malleability functions of the playground to for instance engage schools in reprogramming, and in other ways look into what they call 'pedagogical maintenance'.Some reprogramming has been really small, for instance a tech developer visiting the playground was asked by a child if he could turn the roof light of the hut into a rainbow, which he directly did (and it has remained a rainbow since then).Other reprogramming examples have been more complex, as can be seen in the examples below.

Who Made the Changes
When studying who engaged with the malleability, by making changes to the playground, four main groups were identified: the tech developers, (other) project team members, students, and children.
Most changes were done by the tech developers, the professionals who had developed the tech for the playground.Major usage was related to 'Design after Design', where the design team on several occasions tweaked and changed scenarios, as well as added new scenarios and play possibilities at the playground.On some occasions other project team members also used the scenario programmer to change the playground installations and for instance add new scenarios.This was done by people who are not professional programmers.
A group of around 60 students designed their own scenarios and interactions for the digital playground installations.The created scenarios were then deployed at the playground, some becoming part of the regular scenarios available.According to the course leader (not part of the design team, and from a university not affiliated with the project) the case was popular among the students, especially as they were able to actually activate the scenarios in the playground.Further, student coaches were involved in co-creating scenarios together with children during a Code Summer Camp.
For the everyday malleability function children can for instance make choices by choosing scenarios on the eggs, or changing the story narrative by changing seats in the hut.For active reprogramming children have been involved in two ways.One was via providing input via observations and interviews, meaning that the design team by considering children's actual usages could redesign the playground.A more direct involvement of children in active reprogramming was through the previously mentioned Code Summer Camp.Over two summers 671 children have been engaged; 378 children in 2021, and 293 children 2022.They designed and programmed scenarios and interactions for the playground.The summer camp activities were planned and executed by the project's tech developers.Children's created scenarios and games were then deployed at the playground, some becoming part of the regular scenarios available.

Types of Play-Values Added.
Based on the typologies and taxonomies of play types and play values, changes made to the playground installations connected back to what values the changes enhanced.This was explicated in the design team and discussed in relation to all design decisions.The project used a value-driven design approach [6].Examples of added values are to provide new things to explore at the playground, creating learning situations (for instance learning programming), adding a new type of play interaction (for instance physical or musical), and increased usability.

KEY EXAMPLES OF MALLEABILITY
Below are key examples of use, giving more detailed overviews, while also relating back to the main perspectives discussed in the previous chapter.

Example: The Yearly Calendar for the Hut
Who: Project team member, with help from tech developers, based on observations and interviews with children.What: New everyday malleability: New stories daily.Active reprogramming: Creation of the yearly calendar scenario.Added values: New things to explore each visit to the playground.
Field studies after the inauguration showed several problems concerning interaction in the hut.It seemed difficult for users to understand how to choose within the stories, and the interaction of changing seats broke concentration when trying to listen.It was also hard for a user to see the change of stories over seasons, due to it being such a slow change.After inauguration it was a bit unclear how the malleability of the playground was to be used.For the first month or so it was rather static.The design team continued to have weekly meetings to discuss the playground project.During one of those meetings the issue was raised as to how the malleability possibilities would be utilised.This was in November 2020 and one idea expressed was to create an advent calendar in the Hut, to gain some attention to the playground, and to give a special and changeable experience to children around Christmas.A scenario was designed, using the winter stories, with a small number of story snippets combined to a story episode around 2-3 minutes long.The scenario was designed so that each day a new episode was released.This meant 24 different episodes were released from 1st of December until Christmas eve.The colour scheme of the LEDs in the hut was also changed to Christmas red.A design team member who had never programmed before could, with a lot of help and advice, rather quickly design the base of the scenario.This was the first time the scenario designer was used by someone who was not a developer.
Inspired by the advent calendar, the entire story scenario in the hut was later changed, so that every day would include a new story episode, from the current seasonal stories.The idea was to give the feeling for a child of experiencing something new each visit to the playground.The functionality to choose which character to follow was taken away, since observations and interviews showed huge confusion among users.
The advent calendar became a positive example, for the first time, children could now on a daily basis find new content in the playground.The idea with the advent calendar appealed to the municipality representatives, for the first time showcasing the usage of the re-programmable features, spurring them to further explore the possibilities that this offered.Observations showed that adding the yearly calendar improved the functionality for users, but took away some malleability and interactivity aspects.
"The daughter says "Come, father, into the hut".Dad replies "who do you think it's about today?"And she replies "Maybe the fox!"The father is standing and writing on paper, so I ask her if she has listened to the fairy tales before, she becomes a little shy and stands next to the father but smiles and nods.I ask what she likes the most, "the Rabbit!", she shouts immediately and gets a little shy right after.[...] "Orange first today" says the daughter and looks up and is quiet for a while, listens and says "And Kurran!" [Kurran is the nickname of the Squirrel in the stories: researcher comment] (Translated observation notes from field observations January 2021)

Example: The Musical Eggs
Who: Students, together with tech developers and project team members.What: New everyday malleability: New scenario to choose on The Eggs.Play any music you want.A seasonal scenario activated every Christmas, for playing christmas songs.Active reprogramming: Creation of two musical egg scenarios Added values: Musical play values, learning for the design students.
In collaboration with the local university, 60 design students got an assignment to design scenarios for the playground.They also got to try them on site at the playground, with help from the tech developers.One of the student scenarios turned the egg into an instrument, where each sensor played a different tune, and lit up at the same time.When the scenario started, it played a Christmas song, dots lightning up to show tune order in the song, then users could either just play notes freely on the egg, or try to repeat the song played initially (and if succeeding, getting it played again by the egg, in disco version).The scenario was implemented over the Christmas weekend.After Christmas the design team tweaked the music scenario into a more seasonal neutral children's song, about a small snail.Now, the snail song scenario is activated all year round, except during Christmas when the Christmas song is active.
Using, and building a new version of, the students' scenario raised interesting discussions in the project team.Should the students be reimbursed for their design work?Who owns the scenario ideas developed?Do the scenarios have authors/creators, and is it ok to build on their code to do new scenarios?In the end the solution this time was to contact the students, invite them to a meeting, get their approval for using the scenario, and write an official thank you from the municipality, as well as write a letter of recommendation that they could use for their CV, describing the work they had done pro bono for the municipality, as part of their course.
This malleability example filled many functions.There was learning for the students, and evaluation showed that they appreciated working with an authentic case.The scenarios developed were also a good inspiration for the design team.A new way of interaction with the eggs were developed, implemented, then tweaked further.Later observations showed that the musical egg scenario was easy for users to understand, and popular to play with.

Example: Disco in the Hut
Who: Tech developers together with design team members, based on observations and interviews with children.What: New everyday malleability: Several scenarios to choose from in the hut, drum the stumps to change the lights.Active reprogramming: Creating the scenario disco in the hut Added values: New play opportunities in the form of dancing and drumming.
The yearly calendar took away some interactivity in the hut, and the design team discussed the need to offer some more playful interaction there.Also the team wanted to showcase and try how versatile the system had potential to be, by trying to totally change the experience offered by an installation.Inspired by the popularity of the musical egg scenario the design team decided to make it possible to turn the Story Hut into a Disco Hut.A DJ-interested team member added some music to the scenario designer, and a simple scenario was developed with light effects and music, where the seats were now instead used as drums, changing the lights in the hut when banged upon (or jumped upon).
It proved easier than anticipated to create disco in the hut.At the same time such huge change of functionality and experience also provided new challenges, and some confusion for users, especially related to selection of scenario.Despite this confusion, Disco in the Hut was kept, as observations showed it to be popular.

Example: Scenario Chooser
Who: Tech developers, based on observations and interviews with children.What: New everyday malleability: Making it easier to choose scenarios, and understand how to decide functionality of the eggs and hut.Active reprogramming: Creation of a better scenario chooser.Added values: Easier to choose the type of play you want to engage with.
The field studies showed that it was problematic for users to choose scenarios, and understand how, both on the eggs and in the hut.This became an acute challenge when disco in the hut was created, since there had not been several scenarios in the hut before, no good way to choose between scenarios existed.Therefore, the design team made a tweak, creating a scenario chooser system for both the Eggs and the hut, where they among other things added voices that told users which scenario had just been activated.(E.g, a voice saying "musical egg" when that scenario was chosen).Several other small changes were made concerning for instance colours of scenarios and so on.Field studies showed that this change had a huge impact on improving usability.This example highlights challenges with the malleable playground: since the playground changes, new needs arise (for instance more scenarios available, means more difficulties in how to choose).Changing the actual scenario chooser, not just adding or removing scenarios, provided new challenges, and a risk of jeopardising the whole functionality of the playground.This example also shows opportunities, that with rather small means it was possible to make rather big changes, in a way not possible with a more rigid playground design.The local science park arranges a yearly Code Summer Camp, a programming camp for children.This became an opportunity for the malleable playground features.As part of the Code Summer Camp, the participating children got to learn the programming tool for the playground.Then the children, in groups, got to develop scenarios that were activated in the playground.Around 300 children per year participate in the Code Summer Camp.At the moment 2 scenarios from last year's Code Summer Camp are available at the playground.This means that children who attended the camp can play the scenarios they created, and thus have been an active part in the playground design, using the digital layer.The collaboration also helped the summer camp, offering practical programming challenges, relevant to the children.The playground case has become a recurring one at the Code Summer Camp, and will be run again in the summer 2023.Interesting to note is that all Code Summer Camps involving redesigning the playground have taken place after the end of the actual playground project.Therefore, it shows a possible way to integrate the redesigning into the societal structure, without the initial specific project funding.

DISCUSSION
Building a playground is very expensive, with malleable technology the playground can be changed and updated after inauguration, at a much lower cost than a physical rebuild.It is also a democratic question, to make it more possible for children to affect their environments, and be involved.In traditional playground design some children might be involved for instance as test users, but years later when the playground is finished, those children have already outgrown the playground.Of course, it is also relevant that a malleable technology enhanced playground installation enables tweaks and changes to be made, and adapt it to the users at a specific location or a specific time.This makes the digital layers work like a sort of digital equivalent to lose materials.In the discussion we look at reasons for why the malleability functions were used, tied to the different cases, and connect this to what values the changes aimed to enhance or pursue.Furthermore, in the discussion we look into restrictions affecting the possibility to design a digitally enhanced malleable playground, and suggest future research.

Major Reasons for Using the Digital Malleability Functions
Before inauguration there were many questions regarding how and why the malleability functions of the playground might be used.Looking back at 2.5 years of usages, we can see 5 major reasons to use the malleability functions: 1) Increase usability (without rebuilding): For playgrounds this becomes important since building playgrounds are expensive and time consuming, therefore the possibility to change the playground digitally offers an interesting option to a rebuild.Also, usability when it comes to play can be hard to define, thus requires to a higher extent to actually observe what is going on when children have started to use the playground, and then strive to provide more opportunities for playful usage.

Restrictions When Designing for a Digitally Malleable Playground
The IoT technology enables the malleability, especially it enables programming the interactions from afar.The scenario designer tool is in many ways the main feature for the malleability.During the design process many different ways to approach the malleability was discussed.Some central considerations determined the choices for malleable features, and called for the need to restrict and control how malleability should be used at the playground: 1) Content restrictions: The malleability functions needed to be restricted, in order to prevent the playground from being used for ill causes such as bullying, x-rated content, political messages and so on.This was important for the municipality to avoid such misuse.This also meant that freedom to for instance record sounds or write messages had to be restricted.2) Security restrictions: The malleability functions had to be compatible with the municipalities tech infrastructure, and following their data security policies and such.3) Play interaction restrictions: To not disturb the possibilities to play, it should not be too easy to disturb or destroy other children's play by deploying constant changes.4) Robustness restrictions: Since the project developed a permanent solution, it had to be working from start, and it needed to continue working (unlike a more temporary prototype), thus choosing industry standard solutions and robust options was needed, limiting some malleability options, such as movable parts.5) Resource restrictions: The malleability functions needed to be relatively cheap, as the project had limited budget, and to keep down costs for maintenance.Further, being on site has proved important in order to implement new features, as digital representations cannot fully replace observations in person.
Editing and activating scenarios requires professionals and on-site work, and is a rather resource heavy implementation process, especially compared to the usual low maintenance costs of playgrounds.6) Usability restrictions: In order to be able to involve users, and especially children, the malleability had to be very intuitive, and preferably build on something that children might be able to understand and use.Thus for instance the choice of Blockly, as well as the scenario model, where scenarios could be added or tweaked independent of each other.7) Systematic/policy restrictions: Having a reprogrammable playground is new to the municipality.This poses challenges for instance in finding ways of managing technological maintenance, especially after the project when there is no longer a specific budget for the installations.

Design implications
To summarize experiences from the project, we have gathered our knowledge on designing for malleability in the form of design implications, presented below.This is not to be considered an exhaustive list, but rather some of the most important aspects that can be relevant beyond this specific case, and useful in others design explorations of malleability: 1) Digital layers can offer opportunities for malleability: Tentatively there is potential in using digital layers for malleability in a playground setting, this paper points at some of those.We propose that this approach is worth investigating further.
2) Malleable digital layers as a supplement: We do not propose that malleable digital layers are the salvation of all playground design.Rather, it is to be seen as a compliment to be combined with many other approaches, such as nature play, risky play and other enhancers of play values.3) Give plenty of room for design after design: Make sure you have resources left after inauguration, and use those resources for observations, interviews, design workshops and so on, to be able to make use of the possibility to tweak and improve based on actual play behaviour.4) The installations are not enough: to be able to include children in designing there is a need for not only digital technology, but also collaborations, with for example NGOs and schools.Facilitation and pedagogical tools make it easier for children to co-create.5) System and policy awareness: A connected playground do not exist in a bubble: In a design process, start a dialogue EARLY on how to integrate into existing systems, budgets and policies (or on how to create new ones).This is a major challenge for playground design.6) Intuitive and inclusive onboarding: The playground is for the children, therefore long instructions and the need for parents' help (or borrowing their phones) is to take away agency from the children.Therefore, it is important to focus on how to make the system intuitive and the onboarding easy.7) Appropriation through malleability: To design for malleability has a lot of overlap with designing for appropriation and designing for ambiguity.The combination can provide useful values to a playground.For instance, allowing for interpretation, providing visibility, to let go of some of the control over the interaction, and to invite co-creation.

Limitations
Here, we present a permanent installation of a digitally malleable playground.This permanence had a strong impact on design choices, thereby creating one type of knowledge, but limiting other.While this choice gave certain insights, not least related to design adaptations to rules and regulations around playground design, it has also limited other.In the future, it might be worth exploring similar concepts through temporary prototypes, opening up a broader design space.Further, this is one case with two types of installations, and should not be seen as representative of all cases.Other areas discussed but not implemented include installations in several locations communicating with each other, as well as other sensors and inputs, such as letting current weather data (or direct measurements) affect game functionality.Finally, it would be interesting to further increase the children's agency in the change.Currently, while reprogramming is possible and have been done by children, they need access and support from someone with the rights and knowledge.This would require balancing granting access to direct change with the municipality's need for control of their public spaces.

CONCLUSION
In this work, we set out to introduce digital malleability into a traditional playground design and show how malleability opens up new ways of using and changing the playground, compared to a more traditional rigid playground design.The project showed a possible way of adding a digital malleable layer to a playground by designing IoT enhanced play installations.It proved possible to create digitally malleable playground installations, and to include them permanently into a public outdoor play environment, albeit with some major restrictions to the extent of the malleability, caused by for instance policies, available resources and security.It was found that the malleability functions were used on several occasions, for different reasons, and by different stakeholders, including the children.Tentatively we suggest that digitally malleable installations provided added value to the playground, compared to more rigid playground design, in terms of adding new play values, offering new things to explore on a daily basis and allowing co-creation and participation from children.However, the malleability functions require resources; such as a functioning design team, time and money, not only during the design process but also after inauguration.This must be taken into consideration when planning for a malleable playground.Further to implement those types of malleable solutions requires to take into consideration the complex ecosystem of municipality requirements, regarding for instance security, policy and technological solutions.This points to a need for a policy level systemic change, in order to make playgrounds more malleable.But this project points towards that such digitally malleable playgrounds of the future can be possible, and also possibly desirable; for the stakeholders, for the children, and ultimately for a more playful and appropriable society.

Figure 1 :
Figure 1: The playground, with the two types of digitally enhanced installations (1 hut, 2 eggs)

Figure 2 :
Figure 2: From left: the egg, inside the egg, the hut at night, the hut with exposed technology, and the scenario designer.

5. 1 . 1
Everyday Malleability.Some malleability was built into the playground installations' everyday functions, for instance the possibility for each child to choose which scenarios to activate, thus affecting in the moment what the installation does.Another type of malleability derives from the fact that the playground changes on a daily basis, for instance with new stories in the hut every day, and stories changing with the seasons, providing new things to explore at each visit to the playground, avoiding rigidity.Below is an example from an observation, showing two groups of children, one on each egg, playing very different games by choosing (and interpreting) the scenarios differently: "The red scenario jumps into action.(...)She walks around the egg and sees the next point of light, touches it and continues about 10 more times, all the time calmly and carefully.(...) Finally she stops, looks towards the other kids at the other egg who are playing with the green scenario, they are more fast paced." (Translated observation notes from field observations January 2021) Another example, showing how designed ambiguity make users negotiate what the purpose of the activity should be:

6. 5
Example: Children Creating their own Scenarios Who: Children age 7-17, students, tech developers.What: New everyday malleability: New scenarios developed for The Eggs.Active reprogramming: Children creating their own scenarios.Added values: Co-creation involving children, learning programming for children, feeling of appropriation (children's own scenarios at the playground).
2)Increase play possibilities: adding more play values, adding new types of scenarios, make room for re-exploration of the playground as it changes.3) Involving children: The malleability functions have been used to let children design their own scenarios.Albright with supervision and some help.Thus, involving them in the playground design.Also, findings from observations and interviews with children can be transformed into almost immediate actual changes and tweaks of the playground, not just becoming input for a new planned playground, years into the future.More directly, the digital layer offers a possibility to choose actively what the playground should function as in the moment, thus appropriating it by for instance choosing ones favourite scenario.5) Learning situations: The malleability functions have been used on several occasions, and with promising results, to for instance inspire interest in programming, learning concrete block programming and so on.