Share or Waste? Using an ICT-platform to share food on a university campus Cecilia Katzeff, Annika Carlsson Kanyama Jorge Zapico Dept. of Sustainable development, environmental science Dept. of Computer Science and Media Technology and engineering Linnaeus University KTH Royal Institute of Technology Växjö, Sweden Stockholm, Sweden jorgeluis.zapico@lnu.se ckatzeff@kth.se, annikack@kth.se Abstract— Considering that food production for human Food waste has been defined as food losses occurring at the consumption has a large environmental impact, food waste is end of the food chain (retail and final consumption) and relates major challenge for sustainable development. Although food waste to retailers’ and consumers’ behaviour [7]. For reducing food occur at all phases in the food supply chain, private consumption waste several strategies have been suggested such as awareness has been identified as a major phase of food waste generation. Intervening at this phase provides an opportunity of change. The raising among consumers and retailers exemplified by Quested article reports the testing of a digital prototype designed to et al. [8]. Others have suggested to decrease the plate size at facilitate for employees and students at a university campus to buffets with significant reductions in food waste e.g. [9] and share food. A representative group tested the prototype and efforts have also included to mark up the shelves of fridges with associated food sharing activities for two weeks. At the closing of colour codes to keep track of current stocks and thus stopping the test period they filled in a questionnaire evaluating their foods from being wasted by buying unnecessary groceries [10]. experience. Twenty-three responses were obtained showing that In an overview of different interventions to prevent food twelve people used the prototype for collecting food, whereas nine waste at the consumption stage, Carlsson Kanyama et al. used it for sharing their food. Six people did both. Main reasons concluded that many initiatives and strategies formed to for not collecting food included lack of time, unavailability of shared food in their proximity and inaccessibility of spaces where decrease food waste are not evaluated at all [11]. The authors food was located. Main reasons for not sharing food were that they suggest that effects of ongoing initiatives, such as selling not- lack of food to share, lack of time, and that sharing was possible consumed food from restaurants at a lower price should be without the prototype. General conclusions from the study are that evaluated in the short span as well as in longitudinal studies. people will use a digital service for sharing food in the workplace Reynolds et al. [12] came to similar conclusions. In their if there is a critical mass of users and if an effective organization literature review, the authors found 13 interventions quantifying of sharing and collecting food is provided. food waste reductions. The most effective interventions were those that changed the size or type of plates, showing up to 57% food waste reduction in hospitality environments. Index Terms—Food waste, digital prototype, sharing, workplace, user study. A. Digital interventions towards more sustainable food practices I. INTRODUCTION The sharing of consumer products and other resources is not Food that is being produced but not consumed, so called food new. It is something people have done throughout history. loss and food waste, is a major issue at international and national Habits, structures and organization of sharing vary with types of levels as up to one third of all food is spoiled or squandered resources, cultures and how well established the particular type before it is consumed by people [1], [2]. of sharing is. For instance, the sharing of books is a well- This loss and waste also represent a waste of labour, water, established habit in most countries - privately as well as through energy, land and other inputs that went into producing that food. library services. Sharing of cars and spaces is also something Kummu et al. [3] shows that the production of all lost food that people are used to, through for example, car rentals and corresponds to 24% of those total freshwater resources used in hotels. During the last decade, sharing of cars and spaces has food cultivation, 23% of total global land use for crops, and 23% also been facilitated through digital platforms and services such off the total global use of fertilizers. Food losses and food waste as Uber (cars) and Airbnb (accommodation). According to a also contributes with emissions of greenhouse gases in a time national enquiry, the most common sharing transaction in when mitigation efforts need to be enhanced quickly [4]. Sweden was accommodation [13]. Digitalization is a driver for According to the European Commission food waste alone the increase of sharing services through its possibility of generates about 8% of Global Greenhouse Gas Emissions [5]. upscaling and diffusion. Through digital platforms sharing For an overview of the current literature of how food waste services may be spread to groups of people beyond the closest occurs and can be understood see [6]. circle around the resources being shared. Food is not usually in focus when the sharing of resources is depends on volunteer work to manage food safety within the discussed. It differs in substantial ways from the most shared sharing system. Our own study aimed for simplicity, reducing resources because it is not really shared in the sense that cars, management functions to a minimum. apartments, books or tools are. When these physical objects are The idea of digital platforms facilitating food sharing shared they are used by other people through their respective through the creation of social networks has been explored by function, entailing wear and tear. However, they are not Ganglbauer et al. [18]. They studied a specific community consumed in the sense that their use means that they cease to formed around the website Foodsharing.de in Germany. The exist. Sharing food, on the other hand, involves giving away and platform allows consumers, farmers, organizations and retailers receiving food to own. The sharing occurs in the sense of to offer and collect food. It is also linked to a food-sharing sharing something that you have too much of, because you might Facebook group, where broader community discussions take not be able to consume it before it perishes. place. Through studying discussions in the Facebook group, the Although sharing food differs in important ways from authors observe a vibrant community engaging in practical ways sharing other types of consumer products, it may also be at local levels to exchange food mediated by Foodsharing.de. facilitated by digital platforms. As for other sharing activities a They were also engaged on a global level in discussing values digital platform may facilitate the exchange of products and and motivations for the Foodsharing.de community to evolve. reaching a large group of users. Sharing food may also play a Just like the Olio service mentioned previously, Foodsharing.de significant role for the reduction of food waste as giving away seems to be an initiative enabling the reduction of food waste. food that is left over from a cooked meal may prevent edible Nevertheless, there is not enough evidence to provide us with food from being thrown away. Similarly, giving away fruits and data to show that this is the case. vegetables from the garden may lead to them being eaten instead Although there clearly are a few apps and research of turning to soil. This paper explore the question of how a prototypes aiming for facilitating for people to share their food, digital platform may support the activity of sharing food there was none suiting our specific purposes: to be used at involving a larger group of people beyond the closest university campus with a minimum of management. In this type community. That is, the point of the digital platform would be of workplace a lot of food is handled (lunch, snacks, catering that it could draw the attention of food being given away by etc). Also, in this type of workplace we assumed that people people that usually don’t give and receive food from each other. could trust the food being shared, as they have a collegial It would make matches with groups that usually don’t meet in relationship . University campuses are workplaces for both this type of sharing activity. In this way it would also be students and professionals, they host restaurants and cafés, the instrumental in creating a social network around food sharing. inflow of food is considerable, as generally also is true for food waste. Thus, university campuses present an interesting arena for B. Digital interventions reducing food waste lowering food waste through sharing of leftover foods. This has Quite a few efforts have also been made to develop and test not gone unnoticed, but an example of an effort to lower food digital applications for reducing food waste and some examples waste in a campus comes from Pittsburgh University where an are given below. Farr-Wharton et al. [14] developed a app called PittGrub was developed. PittGrub includes a mobile app called EatChaFood that kept track of what is at home notification system to select users to invite to events that have in the refrigerator cabinet, among other things because a camera leftover food [19]. The study, however, focuses on the regularly takes photos inside the fridge, categorizes all food computational aspect of the app and the results don’t reveal items and color codes them. The same author also developed an anything about how useful it was. The setting in a campus app called FridgePal with the aim of reducing waste by e.g. provides several interesting features for sharing leftover foods: keeping track of the best-before dates [15]. Yalvaç et al. reports lots of students who may be willing to collect free leftover foods about a digital app to be launched called EUPHORIA that will due to their own scarcity of resources, lots of employees that are help people keep track of their food and to cook with others with often ordering catering where there might be leftovers and lots the ingredients they already have at home [16]. In short, efforts of employees who may be willing to share and receive leftover have been made to improve storage at home to avoid food waste. foods with each other or with students because of environmental Another strand of digital applications have focused on concerns and interest in innovation and development. In the long facilitating sharing left-over foods with others as a way to avoid run the food sharing platform is intended to facilitate for people food waste. An app called Leftover Swap, allowing the user to to give food away as well as collect food given by others in a find food that other people want to give away or you can give workplace environment, consisting of 12 000 students and 5 700 away food yourself was evaluated in [14]. Results show that lack employees. Thus, a successful the platform could have a great of trust formed an obstacle for receiving food. This could be impact. overcome if the food were packaged, if the people who give The purpose of the article is to report the design and user away the food were known to the receiver or if anyone else study of a prototype of a digital platform to facilitate for recommended those who give away the food [14]. There is employees and student at a university campus to share food with currently an application available in appstore called Olio with each other. The primary research questions we explore are: 714 000 users where private citizens can give away or receive food and other items [17]. Olio seems to be an easy app to use, but to our knowledge, it has not been scientifically evaluated. Olio also • What are the crucial design requirements for a digital test was sent as an invitation and only available for users food sharing platform to facilitate for food sharing in with an email address from the university domain the workplace? • What are the obstacles for sharing food in the workplace? II. METHOD The research method for exploring the above questions consisted in the design and development of a prototype for a digital food sharing platform; a user study of the prototype; and an evaluation of the use through a survey to the test users. The user study allowed us to explore our research questions and the use of the resulting prototype in the university setting. The design of the prototype followed a user-centred and participatory design methodology, where a functional digital prototype was developed and, then, tested together with the end- users in a user study. The concept of the digital food sharing platform was first tested on a smaller group of potential users - three test users from the research team and three test users external to the team. The test users were asked to perform the basic tasks of booking food and posting information on food they wanted to share. The process of how they carried out the tasks was observed. Then, questions were asked concerning ease of Fig. 1. The start screen of the food sharing prototype use, attractiveness, potential impact, etc. Based on results from interviews made with the test group around the concept, a The prototype is at the moment only a research tool and it functional prototype of a digital food sharing platform was was taken down and the data removed after the test and data designed. The prototype was developed in an iterative process analysis. The source code was made available as open source for by the research team with help of developers. The prototype was possible further development. The start screen is shown in Fig. developed as a mobile website using PHP Hypertext 1 and the interface for the booking page is shown in Fig. 2. preprocessor (an open source general-purpose scripting language), Bootcamp and a MySQL database. The prototype had the following functionality: • Sharing food: The users could post information about food that they wanted to share. This included: o Name and description of the food o Picture (uploaded or taken with the mobile phone directly from the page) o Location: as the prototype is developed for use in a university, the users selected a university department from a dropdown, then they could write a more specific location in a text field o Date of availability o Number of portions available • Booking food: The users could see a list of all available food, sorted either by location (department) or by chronologically (latest first). When selecting a food item users could see detailed information about it, and book one portion. The food sharer got a notification that the food had been booked, and if there were not more portions left, the food item would be removed from the available list Fig. 2. Interface for the booking page of the food sharing prototype • Administration: The users had also access to an administration page where they could edit their personal information, see and edit the food they shared, and see a A. User study list of the food items they had booked We recruited 34 participants from university staff and • Registration and login: The prototype functionality students by sending a mail to colleagues in two different was only open for registered users. Registration for the buildings on the campus and to student representatives . Participants were, then, asked to access the prototype through However, only three out of the nine who shared food thought it the weblink, which was included in a mail. All participants were was clear when their shared food had been picked up. This was informed that the user testing would last for two weeks and that also shown by the open response comments, e.g.: it was OK to just pick up food even if they did not have anything to share. At the end of the first week of the test period the My own food was not collected because they couldn’t find it or participants got a reminder to start the testing. A second forgot? reminder was sent at the end of the second week. 1) Procedure of testing There was also some confusion around where to put food to To facilitate the sharing of food, students were allowed be shared: access to a fridge in a place they usually could not enter. During the two weeks of the testing period, 31 persons created an It wasn’t easy to find a good place for the food I wanted to share. account for the app. Out of these, 19 were active users and 12 You would probably need a well marked place for it, since it did not use the app. A total of 42 portions were shared and 28 of might feel strange to collect food it you’re not sure whether it’s these were booked (ordered). The kind of food shared was for the right place. Maybe the app could suggest a general marking example, leftover lunch food from catering, home baked bread, of food so that you’ll be able to recognize it. fruit from private gardens, and packaged food. 2) User survey Twelve respondents used the app to book and collect food. When the test period was over, participants were asked to fill Eight out of these thought it was easy to understand where to out a questionnaire evaluating their experience with the pick up the food, two were neutral and two did not know. There application and the food sharing activities. Questions demanded seems to have been a general unclarity regarding the physical responses in different forms. These forms varied from location of food shared. Some comments illustrating problems checkboxes, multiple choice boxes or in statements, which the in picking up food were: informants were asked to rate by numbers (1-5) corresponding to how true they were judged to be. All questions had a line for comments in free form. It was a bit confusing that food was left in different places. A solution could be that you only see the food that [is put in Questions were grouped into the following sections: places], to which you have access or that food can only be left in one place. • Role at the university • Use of the app to share/pick up food, including reasons for non-use The wrong place was marked for the food I was going to collect, that’s why it turned out to be hard to collect • How easy/hard it was to use the app • How easy/hard it was to pursue a sharing/pick-up of food I never picked up the food I booked. How are you supposed to get into rooms which have a card reader? Maybe you could have • Obstacles for using the app a central place to put the food, so it would be easier to find your • Improvements of the app way III. RESULTS A total of 23 persons responded to the survey. Out of these, 12 You were able to see who was to collect it [the food], but were employees and 11 were students. Nine respondents had sometimes there happened to be a lot of food in the fridge though used the app for sharing food (39%). Twelve respondents (52%) not in the app, which people seemed to forget to collect had used the app for picking up (i.e. booking) food that somebody else had put up for sharing. Six respondents both To collect food was a bit tricky. Us students only had access to shared and picked up food. the small kitchen on floor 1 and a lot of food was put on other floors, which lead to that you could not pick it up. Then there Crucial design requirements were food, which were supposed to be put on floor 1, and which The usability of the app in terms of ease of use and I could not find anyway. An idea would be to mark out fridges comprehensibility were given high scores. Nevertheless, and normal cupboards where you can collect the food, to make improvements could be made concerning the clarity of it easier to know where to leave it and where to collect it. information on where to pick up food and the notification of when food had been picked up. Obstacles for sharing food not related to design A major obstacles for not sharing food was that the participants Out of the nine respondents who had used the app to share did not have any food to share. Eight participants stated this, food two had some problem in understanding where to put their while four stated that they did not have the time and three that shared food to be picked up. Seven out of the nine people who they could share their food without using the app. The major had shared their food thought they were clearly informed when reason, which respondents stated for not picking up food was somebody had booked the food that they had published. that they did not have time to. Another reason was the misunderstanding that a participant could not pick up food unless they also shared food. Other reasons were that: An obstacle for the food sharing app to spread to other • There was no food to share in the same building at the groups to create a critical mass of users could be that sometimes university where the person’s workplace was located it might be easier to share food without using the app. As one • Students did not have access to other storing places than user put it: one particular kitchen • The booked food had already been removed when the Some types of leftover food (e.g. pastries and biscuits) are also person who had booked it came to pick it up easy to just put on the table in the kitchen and, then, you know • No food seemed interesting enough to book/pick up that everybody are free to have some. As stated above, the most common obstacle that respondents From the perspective of reducing food waste, this is of course saw for sharing food according to the multiple choice question fine because food is shared. But the food in this case would only was that they never had leftover food from home. This is an be available for a limited group of people. For instance, it might interesting result considering the fact that private households in not be available to students who might be those needing it the Sweden waste 45 kg edible food per person and year [20]. most. Reasons why respondents from our survey stated that they did Some informal observations were finally made, which had to not have any food to share might be that they were not do with social aspects of sharing of food. One such norm was a representative for Swedish households or that their own left-over reciprocity in sharing food. That is, some participants assumed food did not live up to their own criteria for what kind of food that if a person shares food they are also allowed to collect food could be shared or not. but a person is not allowed to pick up food if they have nothing In responding to the multiple choice question regarding to share. Since this was not a requirement at all mentioned by the reasons for not collecting food shared by others, five respondents app, users’ assumptions might originate from a social norm picked that they don’t trust the quality of other people’s food and indicating that to receive we also need to give. five that they did not want to receive food from people they did IV. DISCUSSION AND CONCLUSIONS not know. Trust and transparency of food cooked by unknown people was a recurring theme as it also was mentioned in the The overall conclusion from the results of the survey is that freely formulated comments from respondents. Examples are the design of the digital prototype for sharing food in the given below: university workplace worked well. This is based on that the 12 people who used the prototype for collecting food, the 9 who To pick up fresh groceries, packaged food, or food left over from used it for sharing their food, and those who did both could fulfil catering feel unproblematic. But it would be more difficult for their activities without problems regarding the interaction with me to pick up somebody’s personal leftover food if I don’t know the prototype. Since half of the respondents were students and that person. half were employees, we could see a pattern in that employees shared food to a greater extent while the students mostly It became clear to me that if I can’t be sure that ingredients collected food. Another conclusion is that the prototype has to which I don’t tolerate (e.g. pepper, banana, oats and coconut be improved before a full-scale intervention. Yet another milk), it would not be tempting to receive somebody else’s food. conclusion is that for realizing the potential of the app a critical It’s too much hassle to need to ask. On the other hand, it’s very mass of users is needed. In combination of a critical mass an easy if the rescued food would be whole foods, i.e. apples or efficient organization of sharing and collecting food is needed. other unprocessed food. It is important that shared food is actually picked up. Otherwise it might go bad and needs to be thrown out, which stands in conflict with the purpose of saving food. This might also cause It’s hard to share cooked food. Maybe it would be most efficient problems in trusting that sharing food actually leads to a to share food from catering and restaurants. The most wasted reduction in food waste and consequently also might prevent personal food would probably be cooked food. This might not be people from sharing. Thus, some kind of information regarding very attractive to share. when food is picked up given to the person who shares their food will be needed. Another significant obstacle in sharing food was that the Logistics and flow is necessary in order for people to trust activity was not part of a recurring everyday routine and, thus, that the sharing activity is working. This includes information sometimes hard to remember. Respondents commented this regarding when food is picked up and pick-up places to be aspect as: situated close to and accessible to all people sharing and collecting food. During the user testing we discovered that Sometimes it’s hard to remember to bring food to share. I usually students don’t have access to the same areas as employees and give away food (or groceries) that I will not eat to family or were therefore shut out of picking up food that was announced friends. on the prototype. Moreover, designated storage places would facilitate the sharing food as well as labelling the shared food. It’s not part of my habits A second significant aspect for the sharing of food in the organization of food sharing has to be worked out on a detailed workplace to succeed is to provide conditions for the formation level through mapping out the target group, the activity flow of of the habit of sharing. A habit is an routine-like behaviour, sharing and collecting in the spaces where the groups move, and described as an automatic response to cues in the environment; through specifying the needs of the respective groups. This as proceeding with little awareness; and as goal-directed [21]. In organization should, then, be reflected in the design of the digital order to facilitate shaping habits around sharing and collecting food sharing platform. Finally, care should be taken to provide food in the workplace, the implementation and marketing stages opportunities for habits around food sharing to be formed. of the food-sharing app are crucial. These stages hold the opportunity of drawing the users’ attention to the existence of V. FUTURE RESEARCH the app, starting narratives and communities around it and, thus, For each ICT application researchers develop, they need to facilitating for its users to support each other making a routine take a step back and reflect around what is really required, at out of sharing food in the workplace. Users may remind each which level of technology, and what the advantages are of an other to bring food from home, to announce leftover catering ICT based solution compared to a low-tech solution. In the case food in the app, and to spread the message of the food sharing of a closed environment as a workspace or, in the case of this activity on the university campus. Patience in seeing habits form test, a university department, there is already a built community will be needed as the formation of new habits take time. In the and certain amount of peer trust. Examples of non-technological case of food sharing it may even involve deactivating old habits, solutions for food sharing could be for example a common fridge such as packaging food at home and bring it to work instead of where everything inside is for share. In this case we could argue saving it in the fridge with the risk of it going bad. One possible that there are certain advantages: backside of these habits is the possibility of rebound effects, as • Trust: Even in a known environment, eating food from users can use the sharing as a way of reducing food waste guilt, unknown sources may be uninviting. ICT provides avoiding other efforts such as adjusting purchases. These effects traceability of the food shared, which presumably adds needs to be explored in longitudinal studies. trust. The results from using and evaluating the digital app for • Extra layer of information: ICT allows to add extra sharing food at the university point to a possible success in information to the food share, this includes for instance: developing it into a proper commercial app extended to the o An availability date so the end-users know whole campus and not only to a limited test group. Some issues when the food was shared. need, however, to be resolved first regarding responsibility and o Text information, this include for example management of the app. t present, there is no business model for information regarding ingredients which may the app that could generate income from its users, as food is be important for people with allergies, given away for free with no fees. Thus, supporting the app will information about special diets such as require funds from a third party or from the university. vegetarian or vegan, or storytelling about the Managing the app could be done by a third party given that there origin of the food such as sharing apples from are funds to pay for it. Another important task if the app would the garden. be launched for the whole campus would be how to market it for • Notification: ICT allows easy overview of the shared achieving optimal use: both in terms of number of users but also items without having to go to the physical place of the in terms achieving a critical mass of users in separate workplaces food, and push notifications so users can know about in different buildings. Likewise, the issue of where to share food items without actively looking. has to be resolved and an idea could be to place refrigerators • Providing feedback on when food has been picked up. used for sharing foods in premises available to both students and To avoid extra work for the people who collect the staff. Some resources for maintaining such fridges would, then, food, this may be managed through sensors or RFID- be necessary. tags on the food. Lack of trust and transparency is an obstacle for sharing food • Forming a social network around food sharing by, for in the workplace. This came to the surface in the testing of the instance, linking to a Facebook group or similar app and has also been observed in other studies e.g. [15]. The quality of the food as well as its contents could be questioned. A main disadvantage is the need for development and The food might contain ingredients that the participant could be maintenance of the technology, and the exclusion of users who allergic to. Food left over from catering showed to be most do not have access to the needed access devices. It may be popular to collect. relevant to see if the advantages of the ICT solutions are enough To accommodate the best interest of all and to use the full for supporting the extra complexity, and if the same advantages potential of reducing food waste the organization of pick-up could be gained by other simpler means, such using paper cards places also needs to be reflected in the design of the prototype. and so on. These questions are relevant not only for the case of Moreover, information of when food has been booked and food waste but also for other cases in the “sharing economy”, whether it has been picked up must be clear. Sometimes food and comparative studies between high-tech and low-tech shared could serve several people and specifying portions solutions could be an interesting research topic. available as well as following up the availability after users had At last, in order for research on reducing food waste through collected portions would need to be clearer. In sum, the the design of technology intended to intervene with people’s behaviour the outcome of user studies needs to be carefully [9] S. Kallbekken and H. Sælen, “‘ Nudging ’ hotel guests to evaluated. The long-term goal is to reduce food waste on the reduce food waste as a win – win environmental measure,” consumption level and to be able to conclude that this goal has Econ. Lett., vol. 119, no. 3, pp. 325–327, 2013. been reached we need to have a quantification of the amount of [10] G. Farr-Wharton, M. Foth, and J. H.-J. Choi, “Colour coding food wasted during the use of the sharing platform. We also need the fridge to reduce food waste,” Proc. 24th Aust. Comput. a baseline level to compare this result to. Such a baseline could Interact. Conf. - OzCHI ’12, pp. 119–122, 2012. consist of e.g. measurements of food waste prior to the use of the [11] A. Carlsson Kanyma, C. Katzeff, and Å. Svenfelt, “Rädda platform or measurements of food waste in a control group. The maten. Åtgärder för svinnminskande beteendeförändringar hos konsument,” Stockholm, 2018. big challenge in this is to arrive at a clear method for measuring food waste at the consumption level. Most intervention studies [12] C. Reynolds et al., “Consumption-stage food waste reduction interventions – what works and how to design of food waste prevention in households have asked respondents better interventions,” Food Policy, vol. In press, 2019. to estimate the quantity of their own food waste [11]. Since this [13] SOU 2017:26, Delningsekonomi. På användarnas villkor. method does not generate very reliable results, more precise and Betänkande av Utredningen om användarna i objective methods to measure household food waste need to be delningsekonomin. Stockholm, 2017. developed. [14] G. Farr-Wharton, M. Foth, and J. H. Choi, “EatChaFood : Challenging Technology Design to Slice Food Waste Production,” UbiComp’13, pp. 559–562, 2013. ACKNOWLEDGMENT [15] G. Farr-wharton, J. H.-J. Choi, and M. Foth, “Food Talks We would like to thank our test users at the university Back : Exploring the Role of Mobile Applications in campus as well as the internal university grant to support the Reducing Domestic Food Wastage,” Proc. 26th Aust. study. Comput. Interact. Conf., pp. 352–361, 2014. [16] F. Yalvaç, V. Lim, J. Hu, M. Funk, and M. Rauterberg, REFERENCES “Social recipe recommendation to reduce food waste,” Proc. [1] FAO, Food Wastage Footprint. 2013. Ext. Abstr. 32nd Annu. ACM Conf. Hum. factors Comput. [2] T. Searchinger et al., “Agency for International Syst. - CHI EA ’14, pp. 2431–2436, 2014. Development),” Agency Int. Dev., no. December, 2018. [17] “OLIO - The Food Sharing Revolution.” [Online]. [3] M. Kummu, H. de Moel, M. Porkka, S. Siebert, O. Varis, and Available: https://olioex.com/. [Accessed: 06-Jan-2019]. P. J. Ward, “Lost food, wasted resources: Global food supply [18] E. Ganglbauer, G. Fitzpatrick, Ö. Subasi, and F. chain losses and their impacts on freshwater, cropland, and Güldenpfennig, “Think globally, act locally,” in fertiliser use,” Sci. Total Environ., vol. 438, pp. 477–489, Proceedings of the 17th ACM conference on Computer Nov. 2012. supported cooperative work & social computing - CSCW [4] V. Masson-Delmotte et al., “Global warming of 1.5°C An ’14, 2014, pp. 911–921. IPCC Special Report,” 2018. [19] M. Silvis, A. Sicilia, and A. Labrinidis, “PittGrub: A [5] European Commission, “Food Waste - European Frustration-Free System to Reduce Food Waste by Notifying Commission.” [Online]. Available: Hungry College Students,” KDD, 2018. https://ec.europa.eu/food/safety/food_waste_en. [Accessed: [20] Swedish Environmental Protection Agency, “Matavfall i 06-Jan-2019]. Sverige. Uppkomst och behandling. Report nr 8765,” 2016. [6] K. Schanes, K. Dobernig, and B. Gözet, “Food waste matters [21] A. Biel, “Environmental Behaviour: Changing Habits in a - A systematic review of household food waste practices and Social Context,” in Individual and Structural Determinants their policy implications,” J. Clean. Prod., vol. 182, pp. 978– of Environmental Practice, Ashgate, 2003. 991, May 2018. [7] J. Gustavsson, C. Cederberg, U. Sonesson, R. Van Otterdijk, and A. Meybeck, “Global food losses and food waste,” Rome, 2011. [8] T. E. Quested, E. Marsh, D. Stunell, and A. D. Parry, “Spaghetti soup: The complex world of food waste behaviours,” Resour. Conserv. Recycl., vol. 79, pp. 43–51, 2013.