PAssive Tracking of people and things for physical beHavior analysis

Project Information

PATH

Grant agreement ID: 703893

DOI

10.3030/703893

Project closed

EC signature date 25 February 2016

Start date 1 June 2016

End date 31 May 2019

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 244 269,00

EU contribution

€ 244 269,00

Coordinated by

UNIVERSITA DEGLI STUDI DI FERRARA
Italy

Periodic Reporting for period 2 - PATH (PAssive Tracking of people and things for physical beHavior analysis)

Reporting period: 2018-06-01 to 2019-05-31

Summary of the context and overall objectives of the project

There is huge interest in tracking the movement of people and understanding crowd behavior in public spaces, such as the one shown in Fig. 1. Such spaces are often monitored by dedicated hardware or by relying on personal devices. However, there are significant privacy and security issues associated to active monitoring. Fig. 2 illustrates the trade-off of privacy preservation and cost saving for diverse active (e.g. cellular, wireless sensor networks) and passive (e.g. RFID, sensor radar networks) systems.

The main goal of PATH is to define a new paradigm for the behavior analysis of people and things, by tracking their positions and dynamics with minimal implementation cost and maximal privacy preservation. Without relying on any targets’ device, PATH enables the detection, localization, and tracking of targets that do not participate in the localization process with minimum intrusion and avoiding the identification of individuals. This is useful for crowd-based decision making to enable the appropriate allocation of resources at critical times.

The objectives of PATH are: (i) to develop a framework for system design and analysis, as well as to derive fundamental limits; (ii) to devise advanced techniques for detection and tracking based on reflections from the targets and exploitation of signals of opportunity; (iii) to recognize new patterns related to targets dynamics based on low-complexity signal feature extraction; (v) to establish a proof-of-concept for an industrial-driven use case.

PATH's theoretical findings provide performance benchmarks and guidelines for system design as well as for developing efficient algorithms. The algorithms and techniques developed in PATH outperform current SoA techniques and maintain low complexity thanks to a new crowd-centric approach. In short, PATH's results open the way to a new class of passive wireless systems as excellent candidates for physical behavior analysis.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The main results carried out in the project are:
1. Development of a framework for the design of device-free localization systems and derivation of fundamental limits of device-free localization in cluttered environments.
2. Proposal of low-level features for crowd-centric behavior analysis in active and passive radar configuration.
3. Introduction of soft information (SI) algorithms for localization and tracking
4. Validation of the framework and the main algorithms through experimentation

1. During the outgoing phase in the Wireless Information and Network Science Laboratory (WINSLab) at MIT, a general framework for the design of device-free localization systems has been developed, which includes multipath, clutter residual, and the interference among different targets.
During the incoming phase at the University of Ferrara, the main findings have been validated through experimentation. Fundamental limits have been derived to determine the maximum accuracy that is achievable for given technology and environment. This is important to analyze and design device-free systems for behavior analysis, e.g. to quantify a performance gap-to-be-filled and to design tracking algorithms that fill the gap.

2. Current algorithms for device-free detection, counting, and tracking rely on multi-target detection, where a different set of measurements is associated with each detected target, for example, by estimating its position and trajectory. This association-based method, also known as individual-centric, leads to unnecessary complexity when the system is interested in behavior analysis rather than individual-level information. For these reasons, PATH focuses on crowd-centric methods, which infer the number of targets and their behavior directly from the measured data without estimating their locations (see, e.g. Fig. 3). Both active and passive radars have been considered: ultrawide-band (UWB) signals for the active case as they allow high spatial resolution, and OFDM signals-of-opportunity for the passive case for their availability (DVB, DAB, WLAN, 4G, and soon in 5G) and low complex signal processing.

3. While conventional localization and tracking approaches rely on the estimation of single-value metrics, higher levels of accuracy can be obtained by using a new approach that obtains soft information (SI) from functions associated with all possible values that such metrics can take. During the incoming phase of the project, one-stage techniques have been developed, including direct positioning and the new methodology of SI encompassing richer information than single-value estimates. A framework for the design and analysis of systems relying on SI is developed. Such a framework has been used to develop machine learning algorithm for crowd-centric counting and behavior analysis.

The application of PATH’s theoretical findings to a use case has been pursued throughout the entire project. The clutter removal and the machine learning algorithms have been validated through experimentation in an indoor office environment via UWB sensor radars. Results show that the techniques developed in PATH outperform current SoA techniques.

The results of PATH have been published on top-tier international journals and presented in the main international conferences of the field. Within one of such conferences, the researcher co-chaired a workshop dedicated to the advances on network localization and navigation in 2017, 2018, and 2019. In 2016, the researcher received the Paul Baran Young Scholar Award of the Marconi Society, which is intended for outstanding young scientists and engineers who have demonstrated exceptional capability and potential within the information and communication technologies. In 2018, she was elected chair of the Young Scholars of the Marconi Society. PATH's findings were also presented to the general public, through the online magazine of the Laboratory for Information and Decision Systems at MIT as well as the social media channels of the University of Ferrara. The researcher presented her project to the Ph.D. students of the University of Ferrara during the incoming phase.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The use of device-free systems guarantees minimum intrusion and individuals are never identified. PATH takes the idea further and eliminates the need for any specialized hardware by studying also the use of source of opportunity, i.e. transmitters that are already in the monitored environment.
The main findings of PATH represent a strong innovation with respect to the existing state-of-the-art. The introduction of the crowd-centric approaches for device-free behavior analysis is a breakthrough for multitarget tracking, which is particularly suited for application in crowd-sensing and behavior analysis.

During the outgoing phase in the Wireless Information and Network Science Laboratory (WINSLab) at MIT, a
during the incoming phase of the project at the University of Ferrara

In the next years, IOT promises to trigger a new industrial revolution encompassing many diverse technologies, also known as Industry 4.0 or Smart Factory. Industry 4.0 is conceived to start with advanced manufacturing and transcend into other segments including ITS for smart logistics, smart cities, and smart buildings. In this context, the monitoring and tracking of people and things through the exploitation of diverse technologies and infrastructures already in place will enable a number of new applications and services.

Fig. 2 - Diagram of Active and Passive Systems

Fig. 1 - Example Application Scenario: The World Trade Center in NYC

Fig. 3 - Example Scheme of Individual and Crowd-Centric Countingg