21/05/2021 10:30

 Aline Carneiro Viana (INRIA) Motion Intelligence

 Aline Carneiro Viana (INRIA): Motion Intelligence (Talk 1: 25/05, 9h) - SafeCityMap project (Talk 2: 25/05, 9h45)


Abstract (Talk 1): 

Challenges related to energy consumption and computation increasing demand are pushing through an architectural paradigm change. The envisaged scenario is to have network intelligence pushed at the edge, much closer to UEs, where learning, reasoning, and decision making will provide distributed autonomy, replacing thus, the classical centralized structures: Integrating collective intelligence in the network is essential. The natural Internet upgrade into a “network of subnetworks” is thus, the new trend, where “local” (more than “global”) evolution is the key principle to enhance situational awareness and adaptation of edge networks.

Yet, a direct consequence of distributed and in-network computation at the edge is their exposure and sensitivity to user mobility, which may render the location of the used edge computation non-optimal in the long run. Unfortunately, current approaches do not encompass the dynamic and uncertain behavior of mobile users. In fact, they are limited by their reliance upon aggregated static users density associated to access points: The individual is at a disadvantage in aggregated-like approaches. Still, related literature do not rely on the user susceptibility to visit new, unexplored places: Past history of movements of an individual who likes exploring new locations provides a wrong estimation of her next movement. Such problems will be exacerbated by the revolutions entailed by 5G. Here, mobility is expected to be managed on demand: Learning and proactive reasoning are essential missed points in 5G.

This talk aims to get sense of daily mobility patterns and motion preferences of individuals, and to give hints on the knowledge to be used to smartly adapt new Internet services to people digital and dynamic behaviour.  

Im01.jpgim01a.jpg

Abstract (Talk 1): 

Challenges related to energy consumption and computation increasing demand are pushing through an architectural paradigm change. The envisaged scenario is to have network intelligence pushed at the edge, much closer to UEs, where learning, reasoning, and decision making will provide distributed autonomy, replacing thus, the classical centralized structures: Integrating collective intelligence in the network is essential. The natural Internet upgrade into a “network of subnetworks” is thus, the new trend, where “local” (more than “global”) evolution is the key principle to enhance situational awareness and adaptation of edge networks.

Yet, a direct consequence of distributed and in-network computation at the edge is their exposure and sensitivity to user mobility, which may render the location of the used edge computation non-optimal in the long run. Unfortunately, current approaches do not encompass the dynamic and uncertain behavior of mobile users. In fact, they are limited by their reliance upon aggregated static users density associated to access points: The individual is at a disadvantage in aggregated-like approaches. Still, related literature do not rely on the user susceptibility to visit new, unexplored places: Past history of movements of an individual who likes exploring new locations provides a wrong estimation of her next movement. Such problems will be exacerbated by the revolutions entailed by 5G. Here, mobility is expected to be managed on demand: Learning and proactive reasoning are essential missed points in 5G.

This talk aims to get sense of daily mobility patterns and motion preferences of individuals, and to give hints on the knowledge to be used to smartly adapt new Internet services to people digital and dynamic behaviour. 

Im02.jpg im02a.jpg

People mobility and activity patterns are general in nature and similarities emerge in different cities around the world. The analysis of these patterns reveals many interesting properties of human mobility and activity patterns. Indeed, human beings are creatures of habit and various aspects of their movement witih a urban area have a high periodicity. People habits are dictated by family life, hobbies, work obligations, or the location of amenities or services.

While all these properties have been investigated at length, the COVID-19 pandemic highly perturbed our mobility patterns and use of urban spaces. This raises two important questions, addressed in the SafeCityMap project. First of all, we investigate how mobility patterns at an urban scale were affected by the pandemic, and especially by harsh lockdown conditions in Spring 2020. As such, 1st-phase SafeCityMap works toward the tracking of the evolution in space and time of population habits in mobility. Second, we believe that the modeling of such patterns as well as the analysis of how they were impacted by the 1st lockdown can provide useful intuitions on the epidemic spread, such as for COVID-19, on different areas of a city. This latter is related to the observation that high population concentration at certain hours and geographical area intuitively increases the probability of agglomeration and, consequently, the contamination and propagation risks. This is particularly probable in small and highly dense geographical zones. Still this claim needs to be investigated from the epidemiological point of view and constitutes not-validated intuitions, which is left for the 2nd-phase of SafeCityMap.

I will thus present the SafeCityMap's data-driven mobility analytics performed on two large-scale datasets collected in Ile de France region, and more particularly, in Paris departments, before and during the first lockdown. Such analysis corresponds to the results of the first phase of SafeCityMap. Our goal is to quantify (in space and time) two phenomena: (1)~the attendance of and (2)~the visiting flows in different urban areas, both before and during the lockdown, so as to quantify the consequences of mobility restrictions and decisions at a urban scale. We believe SafeCityMap outputs can feed Anti-Covid applications (i.e, spots to avoid in a journey, as already done for pollution peaks) as well as draw the authorities attention to particular areas. Results could also be a valid indication of activity loss, and consequently, of the economic impact (e.g., for activity-labeled areas) such areas suffered during the lockdown. Indirectly, on the basis of our results, public policies could be put in place to better separate the different areas of activity, temporal attendance, and density. 


_________________________________________________________________________________

Short BIO

Aline Carneiro Viana: is a Senior Reseach Director (DR) at Inria, where she leads the team TRiBE team. After a 1-year sabbatical leave at the TKN Group of the TU-Berlin, Germany, she got her habilitation degree from UPMC - Sorbonne Universit\'{e}s, France in 2011. Dr. Viana got her PhD in Computer Science from the UPMC - Sorbonne Universit\'{e}s in 2005. Her research addresses the design of solutions for tactful networking, smart cities, mobile and self-organizing networks with the focus on human behavior analysis. She is a recipient of the French Scientific Excellence award since 2015 and for 6 years now and was nominated in 2016 as one of the "10 women in networking/communications that you should Watch" (1st-year nomination of N2Women community).

She has published more than 95 papers, presented in these fields in top-tier conferences and in important peer-reviewed journals. She has been involved in the organizing committee as well as been a TPC member of major conferences (ACM SenSys, ACM Mobicom, IEEE Infocom, IEEE SECON, IEEE LCN). She is Area Editor of ACM Computer Communication Review (ACM CCR) and member of the editorial Board of Urban Computing Spring book series, and Elsevier Ad Hoc Networks. She has coordinated French and International projects (ANR MITIK and EU CHIST-ERA MACACO, and STIC AmSud UCOOL). She was also the coordinator of the Inria Associate Team EMBRACE (between Inria and 3 Brazilian partners).



Parceiros