Our experiment primarily employed open-source software and hardware that anyone with a similar interest can implement in their own environment. The use of these readily available technological devices are already prevalent in many small-scale design projects at home, in the workplace, schools, or aged care facilities where pot plants are enhanced or augmented with LED lights, sounds, or movement, to encourage people to engage with them through various design strategies of persuasion (Fogg, 2009), affordances (Kaptelinin, 2014), and affective design (Picard, 2000).

Based on a more extensive literature review (Loh et al., 2023), we discovered that in the past 20 years, 72 papers published by the Association for Computing Machinery (ACM) discussed projects that specifically used technological devices to engage people with pot plants—for example, through tangible interactions (Adhitya et al., 2016), augmentations of plant qualities (Kawakami et al., 2010), and providing feedback via ambient interfaces (Kuribayashi and Wakita, 2006). In our design project, we employed these types of digital technologies and design strategies to create an infrastructure to enable building occupants to interact and take care of the office pot plants while being mindful of Wakkary's (2021) note that “technologies and humans need to be viewed together as a meaningful whole rather than separately” (p. 6) because “human entanglement with technology is a mutual becoming, in which humans and technologies influence what each other becomes” (p. 101).

The availability of accessible DIY technology and networked devices to assist with the care of indoor plants has opened up opportunities to take care of a plant collectively with others within a community such as growing food within an office environment (Farr-Wharton, 2019); networking a communal plant watering system where residents could care for plants with others in an apartment block (McDonald, 2018); or Plant Hotels (Wu and Koskinen, 2022). Crosby and Vanni (2022) entangled the public with plants and seeds to activate new alliances and understanding. These studies produced evidence that people are more willing to participate in the care of indoor and outdoor plants when being part of a community.

In the recent COVID lockdowns, the security team of a UK office building took their own initiative to care for all the abandoned pot plants across 12 floors of offices, much to the surprise of the plant owners when they returned back from lockdown (Bakar, 2020). The unsolicited provision of care by others within an office reveals the shared sense of being part of a community where care is willingly enacted for others.

When caring for others or for the office pot plant, one is not just concerned for caring about them, but one is actively taking care of the plant and engaging with it to understand its needs. “Caring involves a notion of doing and intervening” (Puig de la Bellacasa, 2011, p. 89). It is this relational thinking that brings about the active carer instead of passive consumer of “nature” in our office pot plant experiment. In relational ethics, “attentiveness and responsiveness to our commitments to one another, to the earth, and to all living things” is demanded over “resolving ethical dilemmas through good moral reasoning” (Given, 2012, p. 750). By engaging with others in the co-production of care, we not only entangle ourselves with the plants and others in the community, we also revert back to how for example, Australian Indigenous peoples have lived in harmony with nature for millennia (Graham, 2014; Graham and Maloney, 2019). When we “decentre the human as the sole source of activity and…elevate the role of the non-human world from a passive backdrop to human activity…[they become] active contributors to relational action as it unfolds” (Frauenberger, 2019, p. 2:21).

We do not see the act of caring as just a benign act of watering the plants but akin to what Haraway describes as the co-transformation through the care relationship with her dog. “Care is not one way; the cared for co-forms the carer too” (Puig de la Bellacasa, 2017, p. 219). Sustaining a co-existence of human and non-human plant requires active caring, and through the care infrastructuring we designed for our plant-human engagement, we wanted to co-create this relationship with our fellow office occupants, both our human colleagues and the non-human pot plants sharing the same office space. In our project, we realized that relying on voluntary kindness or top-down instructions was not enough to introduce the environment in which humans and plants can entangle. Thus, we began exploring opportunities to augment our office spaces with a technological layer. We now turn to introducing our study's design approach.

Not too different from humans, plants require access to some basic life necessities such as light, water, air, and nutrients. Indoor office spaces can only provide plants limited access to these basic life necessities. For example, many office buildings only offer limited window spaces for potted plants to reside at, thus limiting their access to sunlight. Office spaces are designed also to ensure water does not enter indoor spaces, and we therefore need to ensure plants are watered. With these limitations in mind, the common area of our office floor was chosen for our study, because it is the main communal space for the existing office occupants/potential plant carers who gather for their shared meals, to make coffee, or for informal meetings. Here, we chose four locations for the plant shelves (Figure 3) so that the human occupants can easily see and access them. The first shelf location was chosen due to its proximity to a sunlit window (sun shelves). Because there was only one window in the common area, two other shelf locations were chosen in not-so-sunny but highly accessible and visible positions (shade/rest shelves). One additional shelf was placed next to the kitchen sink for easy access to water (water station) (Figures 4–6).

While the positioning of shelves is important, it is critical to consider how office occupants can be made aware of the plants and to participate in becoming their carers. We have chosen to develop some notification systems for the office occupant passing by, to easily understand what and when they need to activate some simple caring actions.

In the marketplace, one can purchase readily available technical parts such as light and humidity sensors, solenoids, small submersible pumps, timers, etc. to make a fully automated plant watering system. One can also develop a system to remind people to water or move plants by attaching moisture or light sensors to the pot to measure the moisture in the soil or amount of available light. This information can be communicated to humans in the plant's proximity via a LED or screen interface, or to everyone via network devices like a smartphone. The main goal of these projects was to communicate or broadcast plant information by embedding technologies in pots to directly monitor the plants' condition. We took a slightly different approach.

The priority we set was to introduce a physical and technological care infrastructure where both office occupants and plants simply exist in situ, and to establish a symbiotic relationship in an impromptu manner, just like a bee's incidental encounter with a flower. The technological layer should be just enough to notify humans of the plants' conditions and allow them to take simple care actions accordingly. It was also essential that with our designed system, we could easily scale up the care infrastructure without much technological oversight and would be simple enough for anyone to make additional contributions, such as bringing their own potted plants. For this project, we made our own pots to ensure they can be easily identified without any confusion, but it could have been any other pots that people choose to add to our care infrastructure as long as it has a flat base to which a small Radio Frequency Identification (RFID) tag can be attached.

The intention was to make this care infrastructure scalable and more inclusive. By asking the human occupants to move plants to their desired locations, they could become an integral part of this infrastructure. Thus, we embedded all technologies under the shelves rather than in each pot (Figure 7). This decision was crucial to ensure that the pots were free of any technologies that required maintenance, fixing or replacement in order to reduce the maintenance cost and time and to allow the community to more freely add or replace pots without the need to deal with installing devices or technological consequences—other than to add an RFID tag and adding plant information to the database that informed us of each individual plant's needs. This approach allowed anyone interested in bringing in their own favorite potted plants to be added to the system (Figure 8). We explain the technologies below in more detail.

The purpose of the technological layer was to notify humans whether each plant needed more light or water. When the pot was placed on one of the marked spots on a shelf, the RFID reader read the RFID tag attached to the bottom of the pot. The microcontroller (Arduino) received and processed the ID unique to each plant and sent it to the server over the Zigbee wireless network, and the server recorded in the database the time the pot was placed on the shelf. The server then retrieved the plant information from the database that maintained the record of when each pot was placed on or removed from which shelf and returned the total duration the pot had been on to the shelf unit that originally sent the ID. In essence, the server maintained a record of how long each plant had been on the sun shelves (receiving sunlight), the water station (given water), and the rest shelves (no extra light or water received). If the pot was not placed on any of the shelves, the system was not able to locate its whereabouts, but it assumed the plant did not receive sunlight or water until it was placed on a shelf (Figure 9).

When the data returned by the server was received by the shelf-unit, its Arduino processed the information to determine the number of orange and blue LEDs (up to 5 each) to turn on. The orange LEDs indicated the amount of sunlight the plant has received, and the blue LEDs indicated the amount of water it had received. Because the pots did not have embedded sensors, the amount of sunlight and water they received were estimated from the information kept by the database about how long each pot was placed on shelves with particular attributes (i.e., sun, water or rest shelves). When the pot had been on the sun shelf for a predefined period, which was determined by the plant's specific needs, all five orange LEDs turned on. If the duration exceeded the predefined time, the last (5th) orange LED blinked to indicate the plant had been under the sun for too long. If it was brought to the water station, the system assumed the water was given by the person who placed it, and all five blue LEDs turned on. With each predefined period passing without sun or water (again, determined by each plant's needs), the number of orange (or blue) LEDs reduced incrementally until only one was on, followed by the last remaining LED blinking to indicate it required sun or water. It was not an automation system. It required a little help from people to satisfy its needs (Figures 10, 11).

The role of the building occupants was to recognize the LED status and respond with appropriate action because there was no third-party vendor responsible for the plants. They were typically informed through the blinking LEDs and invited to take action by moving the pot to a sun shelf, a resting shelf, or the water station to give water whenever they can (refer to Figure 2 introduced earlier).

As mentioned above, each plant had “predefined periods” that defined how frequently it should receive sunlight and water, and the numbers were kept in the database. The predefined periods were determined by initially using commonly understood gardening requirements for the plant's needs and then we adjusted the numbers until the plant was doing well in the environment. Predictably, some plants became less healthy over time despite our effort with the system and required additional care from the plant doctor (the lead author). Most plants lived well beyond our expectations, but some we found did not thrive in the office environment and had to be returned to their original home.

Our design intention was to focus on the plants and the humans interacting with them in their environment, through a technological layer that informed them when the plants needed water or sunlight. The plant's welfare was entwined with its pot, humans, and the technological systems. Humans are a critical life necessity for our office plants, and plants are crucial co-occupants for reasons discussed earlier. In the next section, we discuss the implications of this symbiotic system of plants, humans, and technology which we deployed in our office for over 1 year through observations and interviews with our human study participants during this period.

Six office occupants (fellow design academics) who resided in the offices of the building volunteered to participate and care for the pot plants in the common area for a period of 8 weeks, and they were briefed on how to read the LED signals in order to move the pots from a sunny to shady spot or vice versa, and also how to water the plants.

Two (20–50 min) semi-structured interviews were conducted to obtain feedback from the participants, at the start and end of the 8-week participation period, that is, a total of 12 interviews. At the first interview, initial instructions were given about how the system worked which were then followed by preliminary questions to establish a baseline understanding of the participant's plant caring knowledge, office plant experiences, and level of interest to engage with shared plant caring within the office. For the exit interview, participants gave feedback on what they did, what they thought about the whole process, and suggestions for what worked, what did not, and what other things they thought might work better in order to take better care of the plants.

A sample of the interview questionnaire is included in the Appendix A. The questions in Appendix A are an indicative list as slight variations occurred depending on the interviewee's responses which sometimes required follow-up questions or further clarification. This method of interviewing aligns with the qualitative nature of this research where our physical design intervention is investigated further through technological means, visual observations, verbal interviews, and data analysis. Our 12 data points (6 × 2 interviews) may be construed as a small sample size, however, we consider it appropriate to our study as “sample size alone is not the only factor at play” (Braun and Clarke, 2021, p. 211). We were interested in getting rich data, “getting different stories” to understand the outcomes as we believe that “meaning is generated through interpretation of, not excavated from, data…” (Braun and Clarke, 2021, p. 201). Using simplified coding methods from thematic analysis, we condensed and distilled interview data to find links, correlations, or emergent themes, following established qualitative analytical methods as “concepts and categories are also a means of establishing relations…between different entities” (Lazar et al., 2017, p. 304). The small sample size enabled us to more deeply research “the material within a spiral of conceptual development rather than across a plane of data” to produce findings that “indicate rather than conclude” (Crouch and McKenzie, 2006, p. 491).

The interviews were recorded, transcribed, and the content analyzed qualitatively to identify the main recurring themes and patterns from the plant carer's experiences. These personal narratives provided insights to understanding the implications of constructing an alternative perspective to the caring of and interacting with office plants. Although we sought responses concerning the care protocol and prototypes, we interviewed the participants not as “test subjects” to a design experiment but considered them as active co-constructors of an evolving plant-human relationship through a newly introduced system of office plant care.

The study participants in our design intervention were the six human co-workers inhabiting the office floor and the 16 non-human plants selected to grow in the individual pots. There were three male and three female participants ranging from 30 to 70 years of age who were full-time teaching/research academics in design with individual offices in the faculty building on campus. Four of the six participants have a good to a high level of familiarity with interactive technologies while the other two have a general working understanding. The characteristics of both plant and human participants are further described in Table 1 where human participants were given the choice to take part in our study while the plants did not have a say in being chosen as participants. However, we endeavored to find out how to enable our plant participants to grow well in our environment in seeking this alternative form of caregiving by the community of office occupants.

The plants in our study were given autonomy to flourish, and humans were the support system for them through the digital infrastructure of our designed technological layer. Human participants enabled the office plant ecosystem required for the plants' survival by caring for them, giving them enough sunlight, water, and nutrients to live well. Kallhoff et al. (2018, p. 110) view care as an ecological virtue where the stronger notion of caring for as opposed to caring about indicates not only “a committed relation of holding somebody/something dear, but also the practical engagement of protecting, helping, fostering the other, of promoting the other's good.” Humans provided the office plant ecosystem for plants to survive indoors to enable them to co-exist with us for mutual benefit.

From the analysis of the data collected from the interviews, we created a Human-Plant Carer Profile and also a Plant Participant Profile to describe the different characteristics of the human and plant participants (Table 2). The characterization is developed with a broad stroke as it is meant to provide a frame to introduce and discuss the emerging patterns of behavior in Section 4.2.

Table 2 describes the breakdown of 6 participants as:

This shows willingness of a majority of participants to take care of the plants if the commitment is not onerous. Through the design of the care system with unobtrusive ambient notification and caring tasks simplified through the technological layer, we discovered that it was responded to by participants whom we think could be our main target audience. This, however, does not mean that the other participant characteristics are to be ignored. From initial observations, we believe that the mix of characteristics is essential for the care infrastructural ecosystem to work—which is closer to a representative of an office setting.

Coupled with the physical intervention, we were interested in finding out how office occupants responded to the design intervention. We analyzed this qualitative data thematically to identify four groups of recurring patterns from the participant interviews. Firstly, the interview text was distilled into meaningful clusters by condensing descriptive whole paragraphs into shorter sentences to understand the underlying meanings. Secondly, this information was then organized into thematic clusters by identifying recurring concepts through repeated reading to ensure the essence of the information given by the participant was preserved. The intention was to use simplified coding methods espoused by proponents of this type of thematic analysis (Vaismoradi et al., 2016; Braun et al., 2019) to suit our small design intervention and identify emerging areas of concern for future research. Table 3 shows a sample matrix of how we extracted the emerging themes to understand how participants responded to the design intervention. This list was refined with repeated reading to establish succinct and relevant themes.

In the following subsections below, we expand on the four recurring themes we distilled from our analysis; that is, of Everyday, Simple System, System with Choice, and Individual and Collective caring and knowing.

Usually when technology is added to the function of an object or service, it is meant to decrease the workload of the worker through increased automation. In our project, we purposely designed our plant caring system to be simple, ambient, and neither automated nor aimed at creating efficiencies or optimisations (Powell, 2021). The intention was to give people an opportunity to directly connect with and care for the plant with just enough digital assistance so as to allow them to maintain agency, thereby being active co-constitutive partners of the plant care system.

The ambient LED light notifications were meant to be unobtrusive, giving participants the choice to act on them or to ignore them. No gamification inducements or repeated phone messaging were employed to entice the participant to enact care for the plant to avoid relegating plants to utilitarian objects by using them as proxies for nature, interfaces for experiences or triggers for actions (Loh et al., 2023). The system was not complicated to use in order to be inclusive and enable participation by anyone. It was also designed so that participants were given the latent ability to exercise their own judgement whether to override the system notification for better care (see Section 4.2.3 where a participant made a personal judgement on the care of the plants when there were conflicting choices given by the digital system). We consider this overriding of our digital system to be a success as we want participants to be an active part of the caring protocol—to notice the plants' welfare, use their knowledge to respond to their needs with the help of the light indicators and thus assisting the plant to thrive.

In discussing the changing framework from machine-centered computing to human-centered computing, Tedre pointed out that we should be asking, “what should be automated” instead of asking “what can be efficiently automated” (Tedre, 2008, p. 48). However, we suggest moving the frame from human-centered efficiency to a plant-inclusive environment to ask, “how much do we automate?” We advocate a simple system where humans and technology can co-produce appropriate care for the living “beings” in the office. More automation or outsourced plant care could distance us from understanding plant needs and serve to homogenize plant care as is currently evidenced in ubiquitous indoor plant species seen in any office around the world irrespective of culture or climate.

By augmenting the pot plant with a simple LED light system to indicate water/sun care requirements, our design does not just address the practical requirements of the plant—it changes our relationship with the plant. Verbeek points out, “the relation between humans and technologies is in fact part of a larger relation, between human beings and their world, in which technologies play a mediating role. What is being designed, then, is not a thing but a human world relation in which practices and experiences take shape” (Verbeek, 2015, p. 28).

Highlighting the plant-human-technology relationship discussed above, the distinction between object and thing assisted us to frame the way we designed care for our office pot plants. An object is viewed as a non-living, non-connected entity existing within a closed loop system while a thing is understood as an assemblage of concerns constituting an open system. In this sense, we borrow from Heidegger (1975) and Latour (2005), that every thing is connected and related to each other, in that “[t]hings are not cut off from human relations, but rather are socio-material “collectives of humans and non-humans” through which “matters of concern” or controversies are handled” (Bannon and Ehn, 2013, p. 57). Thus, by viewing plants as interconnected with our daily practices within our sphere of concern, we made deliberate design choices to the fabrication of our pot plants and shelf: (i) The pot was purposely designed to be transparent to focus our attention away from the pot so that we can see the messiness of the soil and be made aware that the plant is not a decorative object but a living organism—an alive other, and (ii) The shelf was built to waist-height so that plants were closer to eye-level to help us connect with them on par. Additionally, by co-locating them as part of social spaces such as coffee making facilities, we also entwine them with our everyday lives where caring for them becomes part of our daily routine.

Some plants had to be removed and replaced as they were starting to do poorly and succumbing to disease. However, this was noticed by some participants who inquired during their interviews why certain plants were missing. We viewed this as affirmation of people interested in the pot plant as a thing and more than just an object. The pot plant is now an assemblage of plant concerns and human participation, manifesting Latour's (2000, p. 10) comment that “[t]hings do not exist without being full of people.”

Our concept of infrastructuring is related to but different conceptually to Brand's (2018) Infrastructure layer, which is seen as a large type of system—such as education, science or water in the natural landscape—that is used to support the functioning of other layers. Here we expand this concept to infrastructuring. Infrastructure is usually viewed as a complex interrelated system that is not normally made visible—for example, below ground service systems for a city or IT networked support systems for a large organization. A shift in the understanding of infrastructure commenced in the 1990s when Star suggested that infrastructure was “both relational and ecological” and inseparable from the context, in that it is always an assemblage, a complex weaving together of many tasks, resources, politics, etc. (Star, 1999, p. 377).

Infrastructuring soon began to be used in design research to denote “situated practices that connect everyday local practices with the larger, often global technological systems” (Bossen et al., 2014, p. 221). Any artifact, object or system that is designed to be deployed within a community is unlikely to be designed successfully as stand-alone but is one that is “co-constructed with the sociotechnical systems that artifacts get adopted into” (Wong et al., 2020, p. 10)—be it digital or botanical. Thus, the success of sustaining the wellbeing of our pot plants is entangled with the desires and actions of the office occupants. The technology we designed to support plant care in the office was not infrastructure that acted as “a substrate for other actions…to run on top of, but rather an ongoing alignment between contexts” as suggested in Ruhleder and Star (as quoted in Björgvinsson et al., 2010, p. 43). We consciously designed our plant care system to function with and within our office community. Andersen advocates Latour's view that “participants are not stand-alone subjects, but network configurations” (Andersen et al., 2015, p. 258).

Technology becomes entangled with our everyday lives by innovating, adjusting, and supporting our care actions toward the potted plants. A different kind of relationship with the office pot plant and technology has been formed, a new kind of socio-material and socio-technological assemblage that did not previously exist was achieved through the use of these new technological devices and a system that formed the technical infrastructure, along with the people who provided the human infrastructure.

Dourish points out that “the embedding of a range of infrastructures into everyday space shapes our experience of that space and provides a framework through which our encounters with space take on meaning” and affirms, “infrastructure and everyday life as coextensive; accordingly, it encompasses not just technological but also the social and the cultural structures of experience in pervasive-computing settings” (Dourish and Bell, 2007, p. 417).

As we designed our digital and physical support systems for the care of our pot plants, we began to realize that “the focal point of design” was “not the completed technological artifact” but the “successful establishment of its usage in a practice context” (Ludwig et al., 2018, p. 113.2). Recognizing that infrastructure “never stands apart from the people who design, maintain and use it” (Star and Bowker, 2002, p. 231), we designed our plant care system where humans were also the infrastructure for the plants as we were very much connected with the flourishing of the plants in our office. Plants, humans and technology were the co-constituents of our office environment, actively entangled in “collaborative infrastructuring” (Ludwig et al., 2018).

The topics of commons and commoning (Avram et al., 2019; Teli et al., 2020) are entangled in our project of care infrastructuring as the care of the plants is not assigned to any particular person but can be enacted by anyone in the building. The office pot plants are usually viewed as part of what is called “common property” of the building as they “belong” to everyone and no-one in particular. Although they are managed by building operators, the building occupants do not own them but we can benefit from the pot plants, akin to shared common pool resources. Hardin warns that items belonging to the commons can be easily exploited to depletion or privatized through governance by a selected few people, resulting in the “tragedy of the commons” (Hardin, 2009). However, Ostrom points out that the “commons are not open access” but “people can and often do act for the common good or for the good of the resource” (Hess, 2008, p. 8).

When the operational budget was cut, the third-party maintenance costs of the office pot plants for our building were eliminated and so were the plants. Thus, this common resource could no longer benefit the building occupant or be shared with the citizen. We then asked ourselves if this type of service provision can be reconsidered under a different lens of a shared commons where “the commons is not a “thing” that we have access to because we hold a title deed or authorization, but something that is ours because we produce and care for it, because we common” (Bresnihan, 2015, p. 99). By co-creating the type of infrastructure for collaborative plant care, we participated in the sharing and caring of the commons of our shared pot plants wherein we considered the commons more as “a social regime for managing shared resources” rather than an accounting asset, thereby able to care for a more-than-human entity, and forge “a community of shared values and purpose” (Clippinger and Bollier, 2005, p. 263).

The participant community cared for the plants and promoted their wellbeing through provision of water and light when informed by the digital system. By providing care in this more direct responsive way, we believe that the pot plant (artifact in Latourian terms) is being de-objectified as the system conveys sociality and its needs to survive, which in turn, raises awareness of the plant's liveliness as an other living being (Puig de la Bellacasa, 2017, p. 87).

With this way of “caring for plants as a relational practice” (Schörgenhumer, 2018, p. 111), we can begin to move away from our present human-centric view of plants where we mostly “regulate the human relation to plants on the basis of commodity-economic logic” (Marder, 2013, p. 185). By “caring for” and not just “caring about” the pot plants, we go beyond just being aware of the situation. By actively caring for the plants, the responsibility and action of caring enables us (all living beings) to “live in the world as well as possible” (Tronto, 1998, p. 16).

Not only do the participants note the pleasure in seeing the plants do well through their care, but their actions also assist the plants to thrive and flourish. Through the system of care that we have devised in our experiment, we believe that we have met Schörgenhumer's three basic requirements of caring for plants that enable a successful relationship through (i) “reading” a plant's needs, (ii) fostering its living conditions by providing water, light, nutrients, space, etc, and (iii) by protecting its flourishing (Schörgenhumer, 2018, p. 112). We assisted participants to read a plant's needs through our LED light system and programming. The system was also designed to enable participants to easily lift up the pot plant to examine the plant at close range, adjust its water and light supply, and also to be able to look at the soil conditions through the transparent pot.

If we consider plants not just as decorative objects but as co-inhabitants of the same office space we occupy, then—for our mutual wellbeing and flourishing—we as humans need to care for the immobile plants and assist them to flourish, too, with a “good life” (Kallhoff et al., 2018, p. 52). By designing conditions for “care entanglements” (Light and Akama, 2014), we enabled the “enhancement of conditions of plant flourishing,” which frames an approach to plant ethics and an ecological virtue of care that “respect(s) plant life” (Kallhoff, 2018, p. 7).