Plant Guild Composer: A Software System for Sustainability Juliet Norton∗ , Alex J. Stringfellow† , Joseph J. LaViola Jr.‡ , Birgit Penzenstadler∗ and Bill Tomlinson∗ ∗ Department of Informatics, University of California, Irvine Email: julietnorton@gmail.com, {bpenzens,wmt}@uci.edu † Department of Environmental Studies, Rollins College Email: alexjstring@gmail.com ‡ Department of EECS, University of Central Florida Email: jjl@eecs.ucf.edu Abstract—This paper presents the design concept for a soft- with an abundance of resources strive to achieve a similar ware application to aid users in the development of a sustainable simplicity, but are inclined to use technology to achieve it. backyard food and resource system1 . It presents how we gathered and analyzed the requirements for an application that achieves A. Ties to RE4SuSy and Related Work in HCI a balance of user convenience, awareness, and sustainability in the context of creating a domestic plant guild. A plant guild is a One of the objectives for the RE4SuSy workshop series is to community of plants that sustains itself and provides for people’s develop Requirements Engineering (RE) techniques that help essential needs. us design software systems that support sustainable lifestyles. Based on a field study, the authors present requirements for Last year’s contributions looked at how to improve existing building a Plant Guild Composer as one approach for addressing the aforementioned challenge. The requirements have been gen- RE techniques for green software [1]. Past HCI research has eralized because future applications that aide in the construction primarily approached the problem of unsustainable practices of a sustainable human system, which supports some human by outlining improper resource usage and its implications need without compromising the ability to satisfy this and other [3]. The issues associated with transitioning to a simple, human needs in the future nor contributing to environmental sustainable lifestyle remain scarce amongst HCI publications, degradation, may encounter similar design challenges. as it is established as a field of research in [2], and not Index Terms—requirements engineering; sustainability; plant guilds; permaculture; human-computer interaction explicitly mentioned in [3]. In contrast to both RE and HCI existing research, this paper provides an example system that I. I NTRODUCTION supports a sustainable lifestyle and may be used as a case study in the future. Complexity characterizes many of our lifestyles. Empiri- cally speaking, it is a side effect of our options for nearly B. Sustainability and Permaculture every aspect of life, from which career to pursue to what From our understanding, sustainable living can be achieved to eat for dinner. The use and development of technology by managing consumption of extraneous materials, i.e., things is often intended to alleviate some of life’s complexity via we replace and dispose of without regard, and increasing convenience, e.g., finding the exact doll your daughter wants self-sufficiency, i.e., providing for yourself. While Pierce ap- for her birthday online is simpler than calling and driving proaches the topic of extraneous materials by encouraging to local toy stores. However, in effort to make life simpler, reflection of how they are used [11], we are interested in technology and the information provided by it has given us enabling people to live independently of these extraneous more choices and options than ever before. In sorting through materials by providing themselves with materials they need. these complexities, the need and ability to act sustainably is Practices like permaculture (the eco-, human-sustainable de- unnoticed or forgotten. In her visit to Change Islands, Phoebe sign for permanence) advocate for similar ideologies, e.g., Sengers discovered a lifestyle characterized by simplicity apply self-regulation and accept feedback [5]. Although in- due to the severely limited access to technology and exotic dependently ensured food and resource security is an innate resources, and the necessity to work for subsistence [13]. In step in becoming self-sufficient, this transition poses rather many ways this simplicity rendered a more sustainable society intensive, short-term complications. Namely, a great time and arguably improved their quality of life. As observed in the investment is required to learn these methodologies and their first authors ongoing ethnography of transition movements in implementation. Central Florida and Southern California, there are communities In permaculture, a domestic plant guild can foster human independence from extraneous materials. A domestic plant 1 Portions adapted from a previous workshop paper [8]. guild, as for example depicted in Figure 1, is a family of Copyright c 2013 for the individual papers by the papers’ authors. Copying permitted only for private and academic purposes. This volume is plants that can sustain itself and provide people with many of published and copyrighted by its editors. their essential needs (e.g., food, building materials, etc.) [7]. It is an enabler of simple, sustainable living, i.e., reduces cost of II. P LANTS G UILDS living, ecological footprint, and the need for consumer goods. In the natural environment, plants exist in complex mu- tualistic communities. Within these communities, or plant C. Example Domestic Plant Guild Scenario guilds, each plant has a unique impact on the surrounding Picture yourself stepping out into a quiet backyard; your environment. Their impact may make one nutrient or resource property is fenced in with an assortment of sugarcane and more available or remove harmful constituents for neighboring bamboo, protecting your privacy, and your other plants from plants. When considering an assessment of plant guilds, we wind. This same bamboo plant was used to build the bench assume that each plant has needs (i.e., inputs), products and you just sat down on. You gently stir your morning tea with behaviors (i.e., outputs), and intrinsic characteristics (i.e., a sweet cutting of sugarcane, yet another function of your physical traits); all are components of the plant’s functional privacy fence. analysis. Once these plant characteristics are understood, a A brigade of bees and butterflies hover busily around plant guild could be created to act as a semi-closed system, the your native goldenrod, gaillardia, coreopsis, milkweed, and only external inputs coming in from the natural environment. sunflower. Where you would normally have grass, the native When this plant guild is used to provide outputs for human and hardy Gopher Apple gladly covers the ground; you even consumers, it becomes a domestic plant guild. occasionally see a Gopher tortoise creep into your yard and lazily munch on his favorite snack. A. Central Objective of a Plant Guild A large persimmon tree provides shade for your bench as A domestic plant guild is one that is designed to insert a well as a nice little treat for you and the local wildlife. A human component into this naturally closed system. Extrane- passionflower entangles the persimmon tree, while its flower ous products from the plants provide edible, medicinal, and bobs happily around in your tea. useable parts. Thus plant selection requires consideration of You begin to search for salad ingredients, your lunch, both fulfillment of the guild’s and human’s needs. The plants occasionally picking a few blackberries and sparkleberries; used in a guild are predominantly perennial (living more than they taste quite like a blueberry, but half the maintenance. You 2 years) or self-seeding annuals, and if at all possible, native pull a couple of leaves off the French Sorrel, the Moringa, and to the region. If the plant guild is designed and established fennel. To top it off, you snag an avocado and some rosemary. properly, the human role is reduced to harvesting what is You don’t have to water or prune, you just have to eat your needed. harvest. This is an ideal domestic plant guild. B. Principles for Designing Plant Guilds D. Challenges in Plant Guild Design - A Need for good RE To achieve a balanced domestic plant guild, experts ini- The design and construction of a plant guild requires time tially model natural system dynamics (i.e., plant community and expert knowledge, two factors that prevent many from functional analysis), then fit it to domestic constraints. The incorporating one. Therefore, if we can provide support tools guild is designed to be convenient for humans (e.g., co- to make it easier for people to develop and establish domestic location of similar plant types and navigable paths) while plant guilds, then the gap between the idea of sustainable still encouraging natural states of action between species (e.g., environments and their realization can be reduced. We believe coaction, interaction, and inaction) [7], and features plants that that utilizing adequate RE techniques [9], [14] in the devel- support the humans’ desired lifestyle (e.g., berries for jam opment of these software tools is essential in providing easy production). The design of a plant guild is also dependent to use yet powerful interfaces to support domestic plant guild on implementation constraints (work, financial, and material construction. resources). In general, the higher the human demand, the more initial energy required for the guild to reach establishment. In context of each constraint, plants are placed in appropriate Outline geographic locations and functional arrangement. If designed This position paper discusses, in the context of the Plant correctly, the plant guild should not require human support past Guild Composer (PGC), the challenges of designing a software its establishment (i.e., when the trees reach maturity), though application with the intended use of supporting a sustainable minor manipulation may increase its robustness (e.g., seasonal lifestyle. Its focus is to mask the complexities of domestic pruning of nutrient rich plants for mulch and fertilizer). How- plant guild creation through simple interfaces utilizing com- ever, due to a variety of variables, success is not guaranteed. It mon and easily observable information about the environment. is possible that further consultation may be required, especially We believe our topic is suitable for the RE4SuSy Workshop for complex, high-yield guilds. The basic design principles that because (i) it focuses on gathering requirements at the inter- permaculture designers follow are lined out in Figure 2. section between human and environmental sustainability, and Due to the intensive process of creating domestic plant (ii) provides an example system in a new application domain guilds, the knowledge required to properly design and establish for supporting the facilitation of sustainable lifestyles while one is currently only possible by experts. As such, we’ve masking complexity. conceptualized the PGC, a tool to support the development of Fig. 1. An example plant guild by Michaelann Velicky. Source: http://elkinsparkfarm.wordpress.com/2011/09/20/pear-tree-guild-debut/ The key species is the pear tree, of which there are two. Mint, chives, strawberry, raspberry, goumi, kale, and fennel are used culinarily. Various parts of strawberry, clover, mint, scorzonera, fennel, New Jersey tea, and purple coneflower can be used for teas. Plants that provide nutrients to the guild include comfrey, clovers, chives, New Jersey tea, goumi, and wild strawberry. Fennel, chives, and mint are used as pest deterrents. New jersey tea, goumi, raspberry, and strawberry attract wildlife, where these in addition to aster and purple coneflower attract pollinators of fruiting plants. domestic plant guilds which, in turn, will empower a simple, In the Winter of 2013, the first author relocated to Southern sustainable lifestyle. California to compare and contrast plant guild design between the two drastically different locations. III. R EQUIREMENTS G ATHERING IN THE F IELD The first author is actively engaging in a longterm ethnog- A. Interview Series raphy of transition networks, specifically those that practice From expert interviews in Florida, five in total, we have permaculture. During the fall of 2011, the first author began come to the conclusion that the user experience design of her participant observation by completing a 9 week, 72 hour the PGC dictates the tool’s ability to effectively model and Permaculture Design Course (PDC) certified by the Green assist the design of a plant guild. Identifying and placing plants Education Center2 and Simple Living Institute3 . It was in in a functional arrangement is difficult for non-experts. They discussions of learning curves and hardships of transitioning also do not have the skills necessary to acquire environmental to sustainable living that the concept of the PGC came to light. information (e.g., soil type, sun patterns, points of erosion, Through the spring of 2012, she continued her observation at etc.) or the means to implement such a complex design. Due to the Econ Farm in Orlando, Florida, sharpening her knowledge plant guild implementation constraints (e.g., budget and time), of plant guilds and other sustainable human systems. For the experience of creating it, beyond the interaction with the designing the PGC, she consulted the Education Director and PGC, must be considered in the tool’s user experience design then Garden Manager (second author) at the University of Cen- and how it models and simulates a domestic plant guild. tral Florida Arboretum 4 , and three community permaculture experts, all of whom are well versed in plant guilds and the B. Central Requirements by Permaculture Designers cultivation of plants for food, medicine, and other materials. From the results of our ongoing field study, we’ve estab- In Fall 2012 the first and second authors continued the field lished the following set of requirements for the design of research at a private residence in Winter Park, FL, installing a the PGC. We have generalized these requirements because plant guild of tropical fruit trees, shrubs, and ground covers. we feel future applications that aide in the construction of a 2 http://www.greeneducationcenter.com/ sustainable human system (i.e., an environmentally beneficial 3 http://www.simplelivinginstitute.org/ system established with the intent to support some human 4 http://arboretum.ucf.edu/ need) may encounter similar design challenges. potential users who are highly motivated to make the sustain- able lifestyle transition, and a lower-bound may be found by observing those who are resistant. 4) Provide the user with the information they need to implement the design and utilize the system: The application should be incorporated into a complete system construction program, whether it is computerized or teacher-guided, so that the system is installed and utilized. Such a program should take into account the directly related socioeconomic concerns of the user in the application design, an implication for design established in [4]. 5) Maximize both sustainability and convenience: Max- imizing both sustainability of the designed human system and user convenience in designing and implementing the system is key for creating effective software in this domain. The convenience metric is dictated by the amount of work the user is willing to do, and will vary depending on user expertise. Systems that focus only on convenience are fre- quently unsustainable (e.g., plastic plates and utensils so the 1950’s housewife doesn’t have to wash dishes after a party). Conversely, systems that prioritize the sustainability of the designed human system may be too difficult for non-experts to use without an intensive time commitment to learning. 6) Require users to provide easily observable environmental data, but use experts and power users to fill in information gaps: Typical users should provide the system with easily observable information so that they become better acquainted with their environment. Experts and power users (i.e., non- expert self-motivated users) will likely put forth more effort than average users to provide the system with detailed informa- tion. This information should be catalogued and used to infer Fig. 2. Permaculture Design Principles after [5]. necessary environmental information not provided by average users (see [12] for an example) and to optimize future designs created by the sustainable human system design application. 1) Use ecocentric and anthropocentric metrics to qualify These generalized design requirements, intended to enable the designed human system as sustainable: The qualifying simple transitions to a convenient, more sustainable lifestyle, metrics will ensure that humans’ needs are supported in are geared towards small-scale sustainable human systems addition to the plants’. These metrics may include, but are not (i.e., for families or communities versus regions or countries). limited to: benefit and utility to the human, protection of the IV. R EQUIREMENTS D OCUMENTATION native ecosystems, and soil fertility restoration. The designs produced via the application must qualify as sustainable to After the first set of interviews and the first data collection alleviate the problems contributing to environmental detriment. sessions of the ongoing field study, we used an artifact- 2) Identify environmental data required to design the most based approach to consolidate the gathered information. The efficient sustainable system: These requirements include data requirements artifact model is based upon earlier work by [6] only available to experts or acquired by professional technol- and [10]. ogy. Given all the necessary environmental data, the applica- The figures provide an overview of the stakeholder model, tion should enable the user to produce a sustainable human the goal model, the context and system vision, and the system design optimized for the intended environment. constraints. 3) Design to condense the time the user spends with the application relative to the entire process of creating the A. Stakeholder Model sustainable human system: The user’s existing knowledge of The stakeholder model (Fig. 3) depicts the organizational the system and receptiveness of the lifestyle change are key and constraining units like the housing authority and the local factors in the time they’re willing to spend in the design government, the research environment with the university, the process. To determine an appropriate amount of time spent advisor and the developers, as well as the future users and using the system an upper-bound may be found by observing customers for the system. Fig. 3. Overview of the Stakeholder Model D. Constraints The constraints (Fig. 6) provide the most important restric- tions by the rules of the local housing authority that have to be adhered to and influence the design of the plant guilds. These have to be taken into account for design constraints during the modeling of use cases and scenarios. E. Further Development These models and further artifacts will be extended over the duration of the ongoing field study and the design of the system. V. P LANT G UILD C OMPOSER - A V ISION Long before you sat down on the bamboo bench in your Fig. 5. Overview of the Context and System Vision backyard, you faced the challenge of creating the plant guild. When you decided to transition to a simple, sustainable lifestyle you weren’t sure where to start. Then you were B. Goal Model introduced to the Plant Guild Composer, the app that helps you grow a self-sustaining garden. The steep expense of purchasing The goal model (Fig. 4) provides an overview of objectives fresh food, and the deaths of your late potted plants, motivated and goals for the system. There are three types of goals: you to give plant guilds a try. business goals depicted in purple, left side of Fig. 4, usage A. Location Requirements goals in green, top-right of Fig. 4, and system goals in red, on the lower right side of Fig. 4. The application first asked for your address. Once entered, a diagram of your lot containing your house’s location was presented. It then asked you to confirm or edit the diagram. C. Context and System Vision You went outside and walked around the house to make sure the PGC didn’t miss anything, and spotted the utility The context and system vision (Fig. 5) give an overview of connections under the Live Oak canopy in the front yard. the most important elements of the business and operational At this time you also took note of high and low-lying areas, context of the system as well as the core features of the Plant points of erosion, and soil properties. After you completed Guild Composer. your property’s diagram, the program asked for your priorities: Fig. 4. Overview of the Goal Model food, natural medicine, building materials, household items, C. Result: Design Layout wildlife sightings, environmental restoration. You chose food Once the design was finalized, the PGC produced the guild as the primary function and decided that growing building layout and location in your yard, the places to obtain the plants, materials and household items would also be useful on site. tools, and compost within your budget, and instructions on Then you specified cost of implementation in terms of time, how to implement the guild. Two weekends later a mound money, and resources. with thirty young plants setting root was in your backyard. It had only taken you one morning to get the free municipal compost, a day to acquire the plants and about 2 hours to B. Plant Guild Requirements put them in the ground with two friends. You watered the guild a couple times a week at first, then about once a week From there you specified your primary food requirements when they started really growing, until you found they no from a generated list of plants that grow in your climate and longer needed watering. In a few weeks fresh herbs spiced location. First, you investigated the fruits and noticed a native your dinner, in one season the fragrance of flowers was in the variety of persimmon, a fruit you used to buy. When you chose air, and shortly after your first crops were harvested. Now, less to include the persimmon tree, a simulation showed where than a year later, you obtain most of your food, and even some it could be planted on your property. The PGC displayed building materials and cleaning supplies, from your backyard. suggestions for plants, relative to your goals, that could You don’t worry about going to the store in rush hour traffic provide the persimmon tree with its needs. It took into account or minimizing expenses so you can eat healthy. Life really did that your soil was alkaline (pH) and persimmon prefers a become simpler with the domestic plant guild. neutral range. The PGC suggested perennial peanut, a low- VI. D ISCUSSION growing legume that makes nitrogen more available for the persimmon. While you continued choosing plants based upon Our PGC vision is an example of an application designed the PGC’s suggestions, the simulation continued updating in with a balance of user convenience, awareness of user role, size, arrangement, and location. You realized you had wanted and environmental sustainability. to include another plant. It was no longer eligible and the system told you why (too expensive, not compatible with guild A. Reduction of Complexity optimal location), so you removed some plants that were less We believe that the PGC has the potential to reduce the important to add it instead. You proceeded with this until the complexity of transitioning from our modern, complex, con- guild reached a closed system, (i.e., each guild member’s needs sumer lifestyle to one that is simple and sustainable. We is supported by the other members and the environment) and described how the transition to a simple, sustainable lifestyle your desires were met. can be obstructed by up front complexity with the Domestic B. Conveying Sufficient Knowledge Our vision and requirements feature the idea of reduc- ing complexities to the threshold where users still have the opportunity to learn at a higher, more leisurely level. We also believe that by removing all the complexities, the user wouldn’t have the opportunity to understand, utilize, and appreciate the support the system provides (i.e., awareness of user role). We feel that the awareness challenge complements the convenience challenge and that a single solution can be found for both, although it will vary for each sustainable human system. C. Providing Access Accessibility to information is key in achieving the balance of convenience, awareness, and sustainability. It is essential to explore ways to gather information that can’t be acquired from public resources and are too complicated for average users to acquire. Methods that should be researched include: gathering and analyzing information produced by experts and power users into designs, using mobile technologies to aid the user in a more detailed analysis of the environment, and unconventionally utilizing common household items to indicate otherwise elusive properties of the environment. VII. C ONCLUSION In this paper we described the need to develop technologies that enable people to transition into a simple, sustainable lifestyle. We introduced the concept of a domestic plant guild to show how sustainable human systems can effectively support such a lifestyle. We presented a series of requirements for building the Plant Guild Composer, a tool to develop plant guilds. The requirements suggest the incorporation of requirements Fig. 6. Overview of the Constraints analysis techniques to reduce the complexities associated with its creation. Without these techniques, the Plant Guild Composer would only be usable by permaculture and horti- culture experts. We feel that this research area needs further exploration and that our requirements could be applied to tools that support a user’s journey towards a simple, sustainable Plant Guild example. We believe there are many human lifestyle. sustainable systems, especially in permaculture, that also have this complexity challenge. Future Work Designing earthworks for water collection requires intimate Right now we are working on the first prototypes of the knowledge of natural water flow through land. Building an Plant Guild Composer. A screenshot is provided in Figure 7. off-the-grid house requires extensive knowledge of the many ways energy can be produced. We’ve suggested masking ACKNOWLEDGMENTS the complexity by finding the point of greatest convenience that still produces a sustainable system. We believe the RE We would like to thank Kristina Richards at the UCF community should be researching how to enable people to Arboretum for her guidance and support of this work, and utilize sustainable human systems without being bogged down those who provided constructive feedback on early versions by the complexity of learning how to get started; it is a of this paper. contribution RE can provide in the world’s movement towards This material is based in part upon work supported by the simple, sustainable living. National Science Foundation under Grant No. 0644415. Fig. 7. Screenshot of a Prototype of the Plant Guild Composer R EFERENCES [12] J. Ross, N. Shantharam, and B. Tomlinson. Collaborative filtering and carbon footprint calculation. In Sustainable Systems and Technology [1] Richard Berntsson-Svensson, Daniel Berry, Maya Daneva, Jrg Drr, (ISSST), 2010 IEEE International Symposium on, pages 1–6, 2010. Samuel A. Fricker, Andrea Herrmann, Georg Herzwurm, Marjo Kaup- [13] Phoebe Sengers. What i learned on change islands: reflections on it and pinen, Nazim H. Madhavji, Martin Mahaux, Barbara Paech, Birgit Pen- pace of life. interactions, 18(2):40–48, March 2011. zenstadler, Wolfram Pietsch, Camille Salinesi, Kurt Schneider, Norbert [14] Karl E Wiegers. Software requirements. Microsoft press, 2009. Seyff, and Inge van de Weerd, editors. 18th International Working Con- ference on Requirements Engineering: Foundation for Software Quality. Proceedings of the Workshops RE4SuSy, REEW, CreaRE, RePriCo, IWSPM and the Conference Related Empirical Study, Empirical Fair and Doctoral Symposium. University of Disburg-Essen, 2012. [2] E. Blevis and S.C. Morse. SUSTAINABLY OURS: Food, dude. Interactions, 16:58–62, 2009. [3] C. DiSalvo, P. Sengers, and H. Brynjarsdóttir. Mapping the landscape of sustainable HCI. In Proc. CHI 2010. ACM Press, 2010. [4] M. Hakansson, L. Nathan, G. Leshed, Mann S., and E. Blevis. Sustain- ability Community Workshop Proposal. In Proposal CHI, 2012. [5] D. Holmgren. Permaculture: Principles & Pathways Beyond Sustain- ability, 2002. [6] D. Mendez Fernandez, B. Penzenstadler, M. Kuhrmann, and M. Broy. A Meta Model for Artefact-Orientation: Fundamentals and Lessons Learned in Requirements Engineering. In D. Petriu, N. Rouquette, and O. Haugen, editors, MoDELS’10, volume 6395, pages 183–197. Springer, 2010. [7] B. Mollison. Permaculture: A Designers Manual. Tagari, Tyalgum, New South Wales, 1988. [8] Juliet Norton, Alex J. Stringfellow, and Joseph J. LaViola Jr. Domestic plant guilds: A novel application for sustainable hci. Simple, Sustainable Living Workshop at CHI ’12, 2012. [9] Bashar Nuseibeh and Steve Easterbrook. Requirements engineering: a roadmap. In Proceedings of the Conference on The Future of Software Engineering, ICSE ’00, pages 35–46, New York, NY, USA, 2000. ACM. [10] Birgit Penzenstadler and Jonas Eckhardt. A requirements engineer- ing content model for cyber-physical systems. In 2nd International Workshop on Requirements Engineering for Systems of Systems (RESS), Chicago, 2012. [11] J. Pierce. Material awareness: promoting reflection on everyday ma- teriality. In Proc Ext Abstracts CHI, pages 4459–4464. ACM Press, 2009.