Our group presents one Research Paper and four Posters @ICT4S 2020:
The Research Paper “The Unintended Social Consequences of Driverless Mobility Services – How will Taxi Drivers and their Customers Be Affected?” was written by Christina Pakusch, Paul Bossauer and Gunnar Stevens – Social sustainability effects of autonomous vehicles are being discussed in a controversial manner. While there are numerous social benefits to be expected with the advent of autonomous vehicles, some people will face drawbacks of this development. On the one hand for those who earn their living by driving and on the other hand for those customers who are reliant on human assistance to conduct a taxi ride. In order to better anticipate the size of this group and thus how great the threat to the taxi driver’s profession is to be replaced, we analyze secondary data on taxi use in Germany. The results show that the proportion of mobility-impaired passengers is 9%, accounting for about 18% of all taxi trips, indicating that there is a small but significant number of customers whose needs probably cannot be met by autonomous vehicles, and who will potentially continue to rely on human assistance in the future.
The first Poster “Decentralized Shared Mobility – Bringing Peer-to-Peer Carsharing to Rural Areas” written by Paul Bossauer, Lukas Böhm and Christina Pakusch – Carsharing is a type of shared mobility service that has grown in popularity in recent years. Although it is considered an environmental-friendly alternative to individual transport, and ascribed the power to reshape the mobility sector, it is only little prevalent in rural areas. Since the underlying reasons are mostly related to the technical infrastructure, we propose an open, blockchain-based platform that enables peers to share cars in rural areas.
The second Poster “Machine Learning-Based Demand Predictions for Shared Mobility Services” was written by Lukas Böhm, Paul Bossauer and Dennis Lawo – Precise predictions of future demand in shared mobility services could help to facilitate the rapid growth of fleets and support the urgently required mobility transformation. We, therefore, aim to combine recent advances in the deep learning domain to develop a spatial-temporal neural network model to predict shared mobility demand. To build and test our model, we employ real-world data that we collect from numerous shared mobility services around the world.
The third Poster “Really Smart Fridges: Investigating Sustainable Household Storage Practices” written by Margarita Esau, Dennis Lawo and Gunnar Stevens – For a long time now, the ‘smart fridge’ is promised to improve everyday life in private households, supporting healthy eating habits and sustainable food practices. However, current technology is still not widespread and limited in its functions. Similarly, researched prototypes are rather persuasive and not aligned with consumers’ storage practices. We took a practice theoretical lens to investigate current storage practices and actual refrigerators. As follows we present our work in progress and first insights from our contextual inquiry.
The fourth Poster “Supporting Plant-based Diets with Ingredient2Vec” was written by Dennis Lawo, Lukas Böhm and Margarita Esau – Plant-based diets have taken on an entirely new significance as the ecological consequences of diet choice have become more apparent; it is now acknowledged that dietary choices have significant consequences for sustainability. However, plant-based cooking and the ’veganization’ of recipes is something newcomers to these new cuisines struggle with. Attempting to support sustainable food choices and the learning of plant-based cooking, we propose a Word2Vec-based approach for AI-assisted recipe ’veganization’. In this paper, we describe the fundamental thoughts behind the algorithm and explore challenges and opportunities for future research in a short case study.
ICT4S 2020: https://2020.ict4s.org/