Understanding travel behavior is critical for an effective urban planning as well as for enabling various context-aware service provisions to support mobility as a service (MaaS). Both applications rely on the sensor traces generated by travellers’ smartphones. These traces can be used to interpret travel modes, both for generating automated travel diaries as well as for real-time travel mode detection. Current approaches segment a trajectory by certain criteria, e.g., drop in speed. However, these criteria are heuristic, and, thus, existing approaches are subjective and involve significant vagueness and uncertainty in activity transitions in space and time. Also, segmentation approaches are not suited for real time interpretation of open-ended segments, and cannot cope with the frequent gaps in the location traces. In order to address all these challenges a novel, state based bottom-up approach is proposed. This approach assumes a fixed atomic segment of a homogeneous state, instead of an event-based segment, and a progressive iteration until a new state is found. The research investigates how an atomic state-based approach can be developed in such a way that can work in real time, near-real time and offline mode and in different environmental conditions with their varying quality of sensor traces. The results show the proposed bottom-up model outperforms the existing event-based segmentation models in terms of adaptivity, flexibility, accuracy and richness in information delivery pertinent to automated travel behavior interpretation.
Transport mode information is essential for understanding people’s movement behavior and travel demand estimation. Current approaches extract travel information once the travel is complete. Such approaches are limited in terms of generating just-in-time information for a number of mobility based applications, e.g., real time mode specific patronage estimation. In order to detect the transport modalities from GPS trajectories, various machine learning approaches have already been explored. However, the majority of them produce only a single conclusion from a given set of evidences, ignoring the uncertainty of any mode classification. Also, the existing machine learning approaches fall short in explaining their reasoning scheme. In contrast, a fuzzy expert system can explain its reasoning scheme in a human readable format along with a provision of inferring different outcome possibilities, but lacks the adaptivity and learning ability of machine learning. In this paper, a novel hybrid knowledge driven framework is developed by integrating a fuzzy logic and a neural network to complement each other’s limitations. Thus the aim of this paper is to automate the tuning process in order to generate an intelligent hybrid model that can perform effectively in near-real time mode detection using GPS trajectory. Tests demonstrate that a hybrid knowledge driven model works better than a purely knowledge driven model and at per the machine learning models in the context of transport mode detection.
R. Das and S. Winter, “A Neuro-Fuzzy based Hybrid Intelligent Framework for Transport Mode Detection,” in Proceedings of the 6th International Workshop on Mobile Entity
Localization, Tracking and Analysis (MELT), 2016.
Understanding transport mode detection, is important to transport planning and movement behavior analysis. In order to detect the transport modalities from GPS trajectories, various machine learning approaches have already been explored. However, the majority of them produce only a single conclusion from a given set of evidences, ignoring the uncertainty of any mode classication. Also, the existing machine learning approaches fall short in explaining their reasoning scheme. In contrast, a fuzzy expert system can explain its reasoning scheme in a human readable format with more exibility, but lacks the adaptivity and learning ability of machine learning. In this paper, a novel multi-layered hybrid intelligent framework is developed by integrating a fuzzy logic and a machine learning approach to complement each other’s limitations. Thus in this paper an intelligent fuzzy logic-based near-real time mode detection framework is presented that can perform comparable to state-of-the-art machine learning approaches by bridging the trade-off between explanatory power and prediction accuracy of the models. The results demonstrate that the proposed framework has achieved the aim in the context of near-real time transport mode detection.
Human motion trajectories, however captured, provide a rich spatiotemporal data source for human activity recognition, and the rich literature in motion trajectory analysis provides the tools to bridge the gap between this data and its semantic interpretation. But activity is an ambiguous term across research communities. For example, in urban transport research activities are generally characterized around certain locations assuming the opportunities and resources are present in that location, and traveling happens between these locations for activity participation, i.e., travel is not an activity, rather a mean to overcome spatial constraints. In contrast, in human-computer interaction (HCI) research and in computer vision research activities taking place “along the way,” such as “reading on the bus,” are significant for contextualized service provision. Similarly activities at coarser spatial and temporal granularity, e.g., “holidaying in a country,” could be recognized in some context or domain. Thus the context prevalent in the literature does not provide a precise and consistent definition of activity, in particular in differentiation to travel when it comes to motion trajectory analysis. Hence in this paper, a thorough literature review studies activity from different perspectives, and develop a common framework to model and reason human behavior flexibly across contexts. This spatio-temporal framework is conceptualized with a focus on modeling activities hierarchically. Three case studies will illustrate how the semantics of the term activity changes based on scale and context. They provide evidence that the framework holds over different domains. In turn, the framework will help developing various applications and services that are aware of the broad spectrum of the term activity across contexts.
All 5 posters from our group presented at the summit, while the editor was assessing for the award.
We (Kutadinata, Das, Duffield, Jain, Kotagiri, Kulik, Navidikashani, Rigby, Ronald, Thompson, Wang and Winter, with Kelly and Wallace (Monash University)) have won the Best Poster Award at last week’s Disrupting Mobility, a Global Summit Investigating Sustainable Futures held in Cambridge, MA. Our awarded poster, Shared, Autonomous, Connected and Electric Urban Transport, showed results of various aspects of the ongoing ARC Linkage Project Integrating Mobility on Demand in Urban Transport Infrastructures.
Click on the following list to view the presented posters (as PDF files):
Das, R.D., Ronald, N., and Winter. S. (2015), A Simulation Study on Automated Transport Mode Detection in Near-Real Time using a Neural Network. In Proceedings of Research@Locate’15, Brisbane, Australia.
Detecting transport modes in near-real time is important for various context-aware location based services and understanding urban dynamics. In this paper we present a simulated study on detecting transport modes in near-real time using a neural network. We have shown how detection accuracy will vary with different temporal window sizes and different combination of modes. Since in urban environment transport modes move slowly due to traffic, considering movement attributes or kinematics alone for mode detection is not sufficient. That is why we investigated how spatial information can improve mode detection accuracy. The model has achieved 82%-95% accuracy and proves its efficacy over other detection models.
Rahul Deb Das and Stephan Winter will be attending Research@Locate, held in conjunction with Locate 15, starting today in Brisbane, Australia. Rahul will be presenting a paper as part of his PhD research on mode detection from GPS traces. Feel free to talk to them about the iMoD project!
Our current iMoD research students are members of the Graduate Infrastructure Engineering Society (GIES) at the University of Melbourne. Each year GIES organises a conference at which the majority of the 130 research students in the department present their work and produce a poster (students who started recently are exempt). Academic staff members, including Stephan Winter and Nicole Ronald, chair sessions and provide feedback to students.
iMoD presented four presentations during the day: Michael Rigby, Rahul Deb Das and Haifeng Zhao presented in 10 minute slots, while Shubham Jain, as a new student, gave a three minute overview of his planned research. On the day, he was awarded a best presentation prize, based on ratings from both academic staff and research students. Prior to the event, Shubham Jain participated in the Second Slide competition.
We look forward to returning next year with more presentations!
As a Victoria-India Doctoral Scholarship (VIDS) recipient, Rahul Deb Das was invited to present his research as part of the Victorian International Research Scholarships Knowledge Exchange event during Melbourne Knowledge Week 2014. This event was hosted by Victoria’s Lead Scientist, Leonie Walsh, and organised by the Department of State Development, Business and Innovation, who administer VIDS in conjunction with the Australia India Institute.
(c) Mark Avellino
Rahul started his talk with the notion of liveable cities and showed how his research is relevant to improve the quality of liveability of Melbourne and beyond. His research on how smartphone-based travel surveys can help urban planners by providing rich and timely information on urban dynamics for modelling travel demand at different granularities has the potential to improve urban mobility, safety, and various context-aware location-based services.