Abstract: We propose a novel approach for Ad-Hoc WiFi based distribution of information within large crowds of mobile users. The work is motivated by civil protection scenarios where infrastructure based communication often breaks down in cases of emergency. We follow a basic opportunistic networking approach by making use of the smartphones’ built-in WiFi hotspot functionality which in combination with the devices switching between access point and client modes facilitates the propagation of messages on a multi-hop basis. We make three contributions with respect to previous work on this topic. First, we empirically determine core boundary conditions given by the performance of modern smartphones. To maximize system performance under such circumstances we propose novel heuristics for a mode switching strategy based on client mobility instead of random strategies that have mainly been utilized so far. Finally, we compare its performance to a random role switching strategy in a large-scale simulation based on a real dataset consisting of movement traces from 28’000 people during a three day festival in Zurich. Within the simulation we investigate the influence of various parameters on the system’s behavior.

Citation: Franke, T., Negele, S., Kampis, G. and Lukowicz, P. (2015): Leveraging Human Mobility in Smartphone Based Ad-Hoc Information Distribution in Crowd Management Scenarios, submitted to MobiSys 2015.

Download: http://bit.ly/2jcHCpp

Abstract: Collaborative localization is a special case for knowledge fusion where information is exchanged in order to attain improved global and local knowledge. We propose analytical as well as agent based simulation models for pedestrian dead reckoning (PDR) systems in agents collaborating to improve their location estimate by exchanging subjective position information when two agents are detected close to each other. The basis of improvement is the fact that two agents are at approximately the same position when they meet, and this can be used to update local position information. In analytical models we find that the localization error remains asymptotically finite in infinite systems or when there is at least one immobile agent (i.e. an agent with a zero localization error) in the system. In the agent model we tested finite systems under realistic (that is, inexact) meeting conditions and tested localization errors as function of several parameters. We found that a large finite system comprising hundreds of users is capable of collaborative localization with an essentially constant error under various conditions. The presented models can be used for predicting the improvement in localization that can be achieved by a collaboration among several mobile computers. Besides, our results can be considered as first steps toward a more general collaborative (incremental) form of knowledge fusion.

Citation: Kampis, G., Kantelhardt, J.W, Kloch, K., and Lukowicz, P. (2014): Analytical and Simulation Models for Collaborative Localization, J. Computational Science 6 (2015) 1–10.

Download: http://bit.ly/2jcMvP3

Abstract: Collaborative localization is the computation of improved spatial coordinates in mobile agents based on their physical meetings in a pedestrian dead reckoning (PDR) system. Upon meeting the agents can exchange information about their subjective position and update it based on a simple algorithm. We show in a simulation model that the localization error diverges unless this algorithm is introduced in which case it remains bounded. We consider collaborative localization as an example of broader incremental knowledge fusion and discuss its various implications such as the importance of well-informed agents.

Citation: Kampis, G. and Lukowicz, P. (2014): Collaborative Localization as a Paradigm for Incremental Knowledge Fusion, 5th IEEE CogInfoCom 2014 Conference

Download: http://bit.ly/2ilkZKT

Abstract: In this work, we show how a smart-phone worn unobtrusively in a nurses coat pocket can be used to document the patient care activities performed during a regular morning routine. The main contribution is to show how, taking into account certain domain specific boundary conditions, a single sensor node worn in such an (from the sensing point of view) unfavorable location can still recognize complex, sometimes subtle activities. We evaluate our approach in a large real life dataset from day to day hospital operation. In total, 4 runs of patient care per day were collected for 14 days at a geriatric ward and annotated in high detail by following the performing nurses for the entire duration. This amounts to over 800 hours of sensor data including acceleration, gyroscope, compass, wifi and sound annotated with groundtruth at less than 1min resolution.

Citation: Gernot Bahle, Agnes Gruenerbl, Enrico Bignotti, Mattia Zeni, Fausto Giunchiglia and Paul Lukowicz (2014): “Recognizing Hospital Care Activities with a Pocket Worn Smartphone”, 6th International Conference on Mobile Computing, Applications and Services (MobiCASE 2014)

Download: http://bit.ly/2jcB3mo

This work was presented at HAIDM 2015. The 2015 workshop on Human-Agent Interaction Design and Models was co-organised by SmartSociety.

Abstract: We study simulation models of spreading on peer-to-peer communication networks where any peer (or agent) can be the source of information, be it sensory recognition or contextual knowledge. In such a situation the value or quality of information is of key relevance. Questions of trust, provenance and the problem of the interaction pattern arise and are approached by three different algorithms in our paper: (i) “quantitative democracy”, where knowledge is averaged on a meeting (ii) “experience takes all”, where the more experienced (the teacher) overwrites all prior knowledge of the less experienced (the “student”), and (iii) “transitive experience” where not only information but also experience is handed over. We compare these different regimes and identify their tradeoffs.

Keywords: Trust, provenance, self-organization, emergence, collaborative information processing.

Citation: George Kampis and Paul Lukowicz. Collaborative Activity Recognition.

Download: http://bit.ly/1SQJhad

Smart-Watch Life Saver:
Smart-Watch Interactive-Feedback System for Improving Bystander CPR

This potentially life saving technology, uses a smart-watch to unobtrusively guide its user in efficiently delivering CPR, and is SmartSociety’s first step in developing the Smart Nurse project.

The 19th International Symposium on Wearable Computers (ISWC 2015) was a conference dedicated to cutting-edge research in wearable technologies, and took place on 7-11 September in Osaka, Japan.

You can find the paper through our website, here, or in the conference proceedings, here.

Abstract: In this work a Smart-Watch application, that is able to monitor the frequency and depth of Cardiopulmonary Resuscitation (CPR) and provide interactive corrective feedback is described. We have evaluated the system with a total of 41 subjects who had undertaken a single episode of CPR training several years previously. This training was part of a First Aid course for lay people, commonly accessed in this population. The evaluation was conducted by measuring participant CPR competence using the “gold standard” of CPR training, namely frequency and compression depth. The evaluation demonstrated that the Smart Watch feedback system provided a significant improvement in the participant performance. For example, it doubled the number of people who could maintain both the parameters in the recommended range for at least 50% of the time.

doi: http://doi.acm.org/10.1145/2802083.2802086

Citation: Agnes Gruenerbl, Gerald Pirkl, Eloise Monger, Mary Gobbi, and Paul Lukowicz. 2015. Smart-watch life saver: smart-watch interactive-feedback system for improving bystander CPR. In Proceedings of the 2015 ACM International Symposium on Wearable Computers (ISWC ’15). ACM, New York, NY, USA, 19-26.

Citation: http://bit.ly/1NgCIlk

Our paper “Recognizing Hospital Care Activities with a Pocket Worn Smartphone” received the Best Paper Award at the 6th International Conference on Mobile Computing, Applications and Services (MobiCASE 2014).

This has been a joint effort of the German Research Centre for Artificial Intelligence and the University of Trento, and concerns the problem of “semantic gap” between machine and human semantics.

Abstract: In this work, we show how a smart-phone worn unobtrusively in a nurses coat pocket can be used to document the patient care activities performed during a regular morning routine. The main contribution is to show how, taking into account certain domain specific boundary conditions, a single sensor node worn in such an (from the sensing point of view) unfavorable location can still recognize complex, sometimes subtle activities. We evaluate our approach in a large real life dataset from day to day hospital operation. In total, 4 runs of patient care per day were collected for 14 days at a geriatric ward and annotated in high detail by following the performing nurses for the entire duration. This amounts to over 800 hours of sensor data including acceleration, gyroscope, compass, wifi and sound annotated with groundtruth at less than 1min resolution.

Index Terms: Activity Recognition, health care documentation, real-world study

Citation: Gernot Bahle, Agnes Gruenerbl, Enrico Bignotti, Mattia Zeni, Fausto Giunchiglia and Paul Lukowicz (2014): “Recognizing Hospital Care Activities with a Pocket Worn Smartphone”, 6th International Conference on Mobile Computing, Applications and Services (MobiCASE 2014, http://mobicase.org/2014/show/home)

You can find the complete list of proceedings here: http://proceedings.dev.icstweb.eu/2014/mobicase2014/file-storage/index.html#submissions

[wpdm_file id=38]