Estimation of Greenhouse Gas Emission Reduction from Shared Micromobility System

Shared micromobility is widely recognized as an environmentally friendly travel mode and a critical component of transportation decarbonization. However, quantitatively assessing its environmental impact using real-world trip data is an unresolved and challenging subject. To this end, we proposed a system combining machine learning algorithms and Monte Carlo simulation to address this issue. First, several machine learning algorithms (Random Forest, XGBoost, and LightGBM) were utilized to identify citizens’ travel mode choice preferences and then estimate the substituted travel mode of each micromobility trip. Second, to ensure the reliability of the final environmental impact assessment, the Monte Carlo simulations were used to simulate the substituted mode of each trip. Finally, the environmental impacts were calculated based on the life cycle greenhouse gas emissions.


Context Aware Photo Protection for In-Situ Sharing

People nowadays are getting used to using mobile phones for daily photography, and sharing these precious moments online or in-situ with their friends. However, there is a potential risk of privacy leakage during the in-situ photo sharing process. To address this risk, we propose OASIS, a cOntext Aware photo protection for in-SItu sharing behavior: using the front camera to tell different viewer, OASIS customize viewer’s photo gallery seamlessly between different viewers according to the context of the photo. In this way, we can provide viewers with a good sharing experience while protecting the privacy of the owner.

Accelerating Mobile Computer Vision Applications with Edge Computing

Modern mobile applications have become computationally intensive as recent developments in machine learning and computer vision have instigated a new consumer trend for virtual applications. Most of the heavy lifting is traditionally offloaded to cloud servers; however, network latency hinders user experience, making real-time rendering implausible for sophisticated applications. Although mobile devices are capable of performing necessary real-time calculations, mobile devices typically cannot sustain heavy workloads. Edge servers become useful in these scenarios as they are powerful and close enough to users to support real-time applications. Computer vision applications are currently being investigated.

Inferring Social Relations by Using Real-World Sensing Data

People interact frequently with others in their daily life. Mobile phone has become one of essential parts of people’s daily life, data collected from mobile phones have the potential to infer the relational dynamics of individuals. This study addresses the problem of interpre(ng social rela(onships from human – human interactions captured by mobile sensing networks.

Distance Measurement in Diffusion-based Molecular Communication

Molecular communication (MC) has been widely studied recently, because of its feasibility at Nano-scale and bio-friendly character over conventional communication techniques. Since communication distance is fundamental for reliable connection among other related applications, efficient and robust distance measurement in MC is desired. To this end, we investigate an approach using arrival time difference for diffusion-based molecular communication systems.