Contributions published at Informatics and Sustainability Research (Lorenz Hilty)

This literature report has been written as preparation material for the Virtual Conference on University Air Miles Reduction, taking place between October 30th and November 6th, 2017. The Virtual Conference project has been initiated by ETH Zurich and the University of Zurich under the patronage of the IARU network. It is only possible thanks to the great support from the participating universities.

Information and Communication Technology (ICT) is an important enabler for a low-carbon economy in Switzerland. ICT has the potential to avoid up to 3.37 times more greenhouse gas (GHG) emissions than the amount of emissions caused by the production, operation and disposal of ICT devices and infrastructures used in Switzerland in 2025. In absolute terms, ICT will enable the Swiss economy to save up to 6.99 Mt CO2-equivalents (CO2e) per year, with an own carbon footprint of 2.08 Mt CO2e per year. This opportunity for the ICT sector to contribute to climate protection, however, can only be realized under optimistic assumptions. In particular, it is necessary that the existing technological and economic potentials are systematically exploited by taking ambitious and targeted actions. Such actions can be especially effective in the transportation, building and energy sectors, which have the highest potential for ICT-enabled (“smart”) solutions to reduce GHG emissions. At the same time, the carbon footprint of the ICT sector itself must be reduced by 17%, which is technologically and economically feasible due to efficiency gains.

The WOKO is a cooperative which manages over 60 properties in Zurich and Winterthur. They are lending single rooms or studios to students, mostly limited to one year. Therefore every year numerous rooms and studios have to be checked and handed over. In the context of the bachelor thesis the process, called business process change of tenants, is analysed and searched for weaknesses. To optimise the hand-over a to-be process is developed and on its basis an Android tablet application to support the business process change of tenants is designed and implemented.

This thesis compares different workstations at the Department of Banking and Finance, while taking eco-efficiency into account. The current state of research literature suggests, that the size of electronic devices correlates with its economic impact, with smaller devices leaving a smaller ecological footprint. Based on my own collected data, this study shows that an expansion of a thin client solution is worthwile for most users. The biggest positive impact on the ecological impact is generally caused by a extension of the devices' life cycle. To maintain user acceptance and benefit on a high level, individual needs must be considered.

Textmining for information extraction from unstructured information is of growing importance with the permanent increase in computing capacity. In the area of customer review evaluation for products from on-line stores there are some initial research approaches reported. Such aim for feedback on product related properties. The significance of this-like extracted information still lacks sufficient interpretation of human language due to limitations of computers and software. Related research attempts to address these shortcomings by improving the computer’s ability to interpret human language by improving textmining methods. The presented work develops and tests a different approach for bridging the gaps by combining several different existing textmining analyses methods but also enable the user to apply interpretation corrections interactively, which is expected to enable improved product analysis information. To approve this a functional prototype was developed, programmed in the computer language “R”, tested and evaluated. The motivation was to gain valuable insight into such textmining approaches in order to derive ideas for improved method development, which in future could improve practicability for end users and not being restricted to the use of larger enterprises.

Organizations like Administrations, Schools and Businesses produce records during operation. These records contain critical data and need to be processed, administered, and stored in a controlled fashion. To support the processing, administration and storage of records, some form of a records management system (RMS) is used. This paper defines a physical RMS (records are in paper form) and a digital RMS (records are in PDF or similar form) for a simulated small to medium sized swiss businesses and compares the two systems regarding energy usage and space utilization. In addition to the comparison based on the simulated business an Excel-tool was developed to be able to change the underlying assumptions used for the comparison via a set of parameters.

Life Cycle Assessment (LCA), the most prominent technique for the assessment of environmental impacts of products, typically operates on the basis of average meteorological and ecological conditions of whole countries or large regions. This limits the representativeness and accuracy of LCA, particularly in the field of agriculture. The production processes associated with agricultural commodities are characterized by high spatial sensitivity as both inputs (e.g. mineral and organic fertilizers) and the accompanying release of emissions into soil, air and water (e.g. nitrate, dinitrogen monoxide, or phosphate emissions) are largely determined by microspatial environmental parameters (precipitation, soil properties, slope, etc.) and therefore highly context dependent. This spatial variability is vastly ignored under the “unit world” assumption inherent to LCA. In this paper, we present a new calculation framework for regionalized life cycle inventory modeling that aims to overcome this inherent limitation. The framework allows an automated, site-specific generation and assessment of regionalized unit process datasets. We demonstrate the framework in a case study on rapeseed cultivation in Germany. The results from the research are (i) a better understanding of the characteristics and application of spatial data in the context of LCI modeling, (ii) a framework for generating regionalized data structures, and (iii) a first examination of the characteristics and significance of further use cases.

L’aspiration à mettre en œuvre des politiques basées sur des faits tend à encourager l’utilisation d’indicateurs et d’objectifs quantitatifs définis à l’échelle européenne pour élaborer des mesures et pour évaluer leur impact. Les auteurs avancent que cela réduit le champ des discussions et empêche la création d’une stratégie européenne volontariste pour un développement durable des régions de montagne. Les territoires de montagne ont des caractéristiques sociales, économiques et physiques très variables ; leurs points communs en matière de fragilité écologique, de défis de développement économique et d’exposition aux risques naturels ne sont pas pris en compte par les données conventionnelles. Deux outils complémentaires permettant de compiler et de présenter les faits pertinents pour une gestion durable des ressources sont proposés : le modèle Nexus et la matrice de durabilité (Sustainability Profile Matrix). Ces deux outils reposent sur la constitution d’une base factuelle qui reflète les caractéristiques socio-économiques, les potentiels et les défis de chaque localité ou région. Cependant, ils permettent aussi la compilation d’informations aux niveaux transnational et européen. Cela doit permettre aux administrations locales, nationales et transnationales d’utiliser la diversité territoriale comme levier dans leurs politiques, dans le cadre d’une gouvernance multiniveau des ressources humaines, économiques et naturelles.

The aspiration to implement evidence-based policies has led to an increased focus on quantitative indicators and targets defined at the European level as instruments for designing policy measures and assessing their impact. The authors argue that this constrains debate and has hindered the elaboration of a proactive European strategy for sustainable development in mountain regions. Mountain territories have highly diverse social, economic and physical characteristics; their shared traits in terms of ecological fragility, economic development challenges and exposure to natural hazards are not reflected in mainstream datasets. Two complementary instruments are proposed to produce and present evidence for sustainable resource management and processes: the Nexus Model and the Sustainability Profile Matrix. Both tools entail using evidence that is adapted to the social and economic characteristics, potentials and challenges of each locality or region. At the same time, they make compilations of evidence at the transnational and European levels possible. The objective is to enable local, national and transnational authorities to use territorial diversity as a lever in their policies, within multilevel governance in human, economic and natural resource management. L’aspiration à mettre en œuvre des politiques basées sur des faits tend à encourager l’utilisation d’indicateurs et d’objectifs quantitatifs définis à l’échelle européenne pour élaborer des mesures et pour évaluer leur impact. Les auteurs avancent que cela réduit le champ des discussions et empêche la création d’une stratégie européenne volontariste pour un développement durable des régions de montagne. Les territoires de montagne ont des caractéristiques sociales, économiques et physiques très variables ; leurs points communs en matière de fragilité écologique, de défis de développement économique et d’exposition aux risques naturels ne sont pas pris en compte par les données conventionnelles. Deux outils complémentaires permettant de compiler et de présenter les faits pertinents pour une gestion durable des ressources sont proposés : le modèle Nexus et la matrice de durabilité (Sustainability Profile Matrix). Ces deux outils reposent sur la constitution d’une base factuelle qui reflète les caractéristiques socio-économiques, les potentiels et les défis de chaque localité ou région. Cependant, ils permettent aussi la compilation d’informations aux niveaux transnational et européen. Cela doit permettre aux administrations locales, nationales et transnationales d’utiliser la diversité territoriale comme levier dans leurs politiques, dans le cadre d’une gouvernance multiniveau des ressources humaines, économiques et naturelles.

With the technical development, Smart Home systems have become suitable for daily, mass use. In private households automated systems undertake the task of energy management . These systems require data to realize optimal energy management. The handling of the data is subject of the policies of data protection, which takes action with the purpose of the data. This Bachelor thesis shows the possible applications of energy management in a Smart Homes. It also discusses which data are processed, transferred and collected for the implementation of these possibilities. An experiment is also carried out to determine whether the users are aware of the data collected about them. These data are considered in connection with data protection.

The Master Thesis presents an analysis of interest and requirements towards sustainability at the University of Z¸rich, and proposes a design for a collaborative platform that will support the transition towards sustainability at the university. Through the literature review, questionnaire, and focus group study, the Master Thesis gathers critical aspects of an effective platform, such as personal interaction, informativeness, interactiveness, feedback, relevance, and reach. Wireframes and user stories demonstrate the proposed design for a 'Sustainability Photo Board' platform that encompasses these aspects and create the outline of user interactions. In the design, photo posts and embedded comment threads are supported to discuss sustainability topics. Finally, a functional prototype was created which can run on a local machine to demonstrate all important functions of the proposed system: creating photo posts, commenting, browsing, rating, and searching. The proposed platform addresses concerns to take action, raise awareness, and to share ideas regarding sustainability. The platform aspires to drive involvement and individual commitment in a sustainable campus life by focusing on sustainability issues and on solving them. It is collaborative by nature, as people can discuss issues, rate their relevance, and can even coordinate efforts to address them together. The platform brings personal discussions concerning sustainability online and supports the daily discourse of UZH faculty, staff, and students.

With many parties involved in science, scientific results can be influenced by personal interests. Especially with more and more companies investing in universities and scientific research, the interests of funding entities start to grow in importance. Because of this it becomes easier for companies to follow their commercial interests by influencing scientific results. This can lead to biased results, which can harm the trust the public has in science. To prevent that from happening, transparency about the nature of funding is important. In this thesis it is shown where and in what form funding data can be found and methods to extracting funding information from a paper is proposed and discussed. The two developed approaches use regular expressions and named entity recognition respectively to extract funding entities. Already with a small amount of training data the named entity recognition algorithm performed better than the developed regular expression. The extracted and tagged results are saved in an XML file to be used in further computations.

Excessive salt intake is a global health concern, since 2.5 million deaths per year are preventable if salt consumption was reduced to healthy levels. In Switzerland, salt intake of 94% of men and 77% of women ranges above the WHO recommen-dation of 5 g salt per day. To identify people with an increased salt intake, cumber-some 24-hours-urine-samples or paper-based checklists are used today. This thesis provides the technological and statistical foundations to automatically determine a personís salt intake level from his loyalty card logs. To reach this goal, a system of several software services has been developed during the thesis to obtain the loyalty card logs and measure the daily salt intake of a person by using a food-record-checklist which has been validated and calibrated with 24-hours-urine-samples. The paper-based checklist has been translated into a mobile application running on iOS and Android. Because of the small number of participating users in the thesis study, a correlation between the self-reported salt intake in the mobile app and the loyalty card logs could not be validated yet. The application of classifi-cation algorithms on the created models and the collected data however yielded very promising results. Given a larger sample size and the technical framework developed during the thesis, the correlation between loyalty card logs and self-reported dietary salt intake is expected to be proven in the near future.

This bachelor thesis proposes an improved method for the integration and administration of food related data at the organisation Eaternity AG. To accommodate a set of requirements that have been raised during a workshop with employees at Eaternity, the current data structure was modelled explicitly and in line with a set of key criteria for good data modelling an improved model was crafted. To show the value and functionality of the new data model, a prototype for data capture and administration was built on top of a relational database which implements the model. The prototype implements an architecture which allow for a high reusability of the single components for either a web based or desktop based release. The resulting prototype serves as a template to develop a productive system and the raised requirements introduce ideas for additional projects. The thesis describes the different phases of the innovation process: The observation of the concrete problem, the abstraction of the problem, the definition of goals and the concrete implementation.

The research facility Empa in Dübendorf is working on a study that is exploring the economics of Power to Gas / Hydrocarbon technologies. These technologies convert electricity into methane gas, which can be sold in the high-priced market mobility. The income can be used to pay infrastructure costs of renewable energy. In this thesis the author of the study was being supported. For this reason, a data model was developed with which the collected data was stored in a database. Based on that, a visualization application was developed which can be used to graphically process the data. The application was implemented as an interactive webapplication.

Product lifetimes and disposal pathways are essential aspects of dynamic material flow analyses (MFAs), which have often been used to model stocks and flows of electronic devices and the resources they contain. Existing studies mainly focus on the use and disposal of electronic devices by private consumers. The specific handling of business devices has rarely been assessed. This article presents the results of a study conducted in Switzerland in 2015, comprising interviews with 28 companies. Devices included are desktop and laptop and computers, mobile phones, monitors, televisions, external hard disk drives and servers. Results are compared to data collected from Swiss private consumers. The service lifetime and disposal pathways are fed into a dynamic MFA model to calculate the stocks and flows of business devices. With the example of indium, neodymium and gold, the material resources contained in these stocks and flows are illustrated.

We provide a critique of the development in organisation studies of the idea of ‘unlearning’ as allegedly imported from the psychology literature by Hedberg and understood to mean the manageable discard of knowledge precedent to and aiding later learning. We re-review the psychology literature and in contrast to Hedberg, find that this definition of unlearning is not empirically warranted. We re-examine a selection of highly cited articles in the organisational literature that claim to have conducted empirical research into the Hedberg model of unlearning. We find none provide evidence of its existence. Typically, under the label ‘unlearning’ evidence is provided of a conventional process of theory-change, the setting aside (not deletion) of an established understanding in favour of new understanding when presented with perceived new facts. In all cases that we examine, clear alternative and less problematic concepts should provide a better conceptual framework for the research, such as learning, theory-change, discard of practice and extinction. It follows that the unlearning literature is not in fact the independent, scholarly and scientific literature that many of its adherents believe it to be. We recommend that for concepts allegedly imported from other disciplines more frequent commissioning of cross-disciplinary reviews may encourage the critical works so obviously lacking in the unlearning literature.

Current smart thermostats use a simple strategy when aiming to increase the efficiency of heating and cooling systems, they adjust the temperature whenever the conditioned zone becomes empty. However, while targeting energy savings, these systems often neglect inhabitants’ thermal comfort. We propose to increase thermal comfort by automatically monitoring the inhabitants’ satisfaction with the thermal environment using different hardware components. To this end, we designed the Comfstat infrastructure, capable of collecting essential parameters for thermal comfort prediction. We then use our system to collect detailed temperature and heart-rate data of five users and show that thermal comfort can be inferred automatically from a combination of sensor data within 0.5 points on the ASHRAE scale.

Sustainability is in everyones mind nowadays and makes a deeper understanding of the environmental impact of an institution, company, city or region essential. Environmental management is an attempt to achieve this by learning about the mass flows in the system. In order to reach this target it uses tools like the mass flow network as an aid. The university of Zurich as an institution and its mass flows is subject of interest in this thesis. To assess the environmental impact of these flows, a network must be designed first by using the modelling tool “Umberto“. In a next step by using techniques known from the life cycle assessment the environmental impact can be evaluated. In a final step scenarios are used to identify the sensitivity of certain processes to analyze where future actions would hold the biggest potential for optimization.