Trajetrias de objetos mveis voc j pensou que

Trajetórias de objetos móveis: você já pensou que pode estar sendo monitoriado ? Vania Bogorny vania@inf. ufsc. br

Roteiro n O que são Trajetórias de Objetos Móveis? n Para que servem Trajetórias? n Aplicacões n Problemas com estes dados n Bancos de Dados de Trajetórias n Mineração de Trajetórias 2

A Explosão da Rede Sem Fio Você utiliza algum desses dispositivos ? Você alguma vez já se sentiu monitorado? 3

A Explosão da Rede Sem Fio • Mundo está cada vez mais móvel. . Dispositivos móveis deixam traços digitais que podem ser coletados como trajetórias, descrevendo a mobilidade de seus usuários Geram um novo tipo de dado, chamado “ Trajetorias de Objetos Moveis” 4 4

Exemplos de Trajetórias de GPS: Barcos de Pesca de Atum 5

Exemplos de Trajetórias de GPS: Barcos de Pesca 6

Exemplos de Trajetórias de GPS: Veículos 7

Exemplos de Trajetórias de GPS: Veículos 8

Trajetórias Geradas por Telefone Celular = célula (abrangência de uma antena de telefonia celular) 9

Mobility Data Analysis Several analysis may be done over trajectories: n How people move around the town q n n During the day, during the week, etc. Are there typical movement behaviours? In a certain area at a certain time? How are people movement habits changing in this area in last decade-year-month-day? n Are there relations between movements of two areas? n . . . of 90

Serviços de Localização (Passado) n Limitados a sinais de tráfego fixos 11

Serviços de Localização (Hoje) n Tráfego n Quantos carros estão na Estrada X? n Qual é o tempo estimado para chegar ao destino? n Busca baseada em localização: n n n Quais são os restaurantes no raio de 5 KM da minha posição atual? Onde está a churrascaria mais próxima? Avisos: n Envie cupons a todos os clientes num raio de 4 KM da minha loja 12

Mobility Data Analysis: Applications Trajectory data analysis may be useful in several application domains Veicule Monitoring Transportation Companies monitor their trucks Insurance companies use GPS devices to monitor insured vehicles to reduce insurance price Traffic Analysis To alert people about traffic jams, accidents, etc. . . Identify/predict low traffic regions in a city Security To identify a call

Mobility Data Analysis Animal Migration / Behaviour Analysis Which are the trajectories of migration bird? Where do birds stop? For how long? a given Which is the migration pattern of certain species? Fishing Analysis and Control Are boats really fishing allowed areas? Can we classify vessel trajectories?

Mobility Data Analysis Weather prediction and movement analysis Hurricane tracking

Como é um dado de trajetória computacionalmente falando? Trajetórias brutas: <tid, (x 1, y 1, t 1), (x 2, y 2, t 2), (x 3, y 3, t 3), . . . (xn, yn, tn)> 16

Exemplo de uma tabela com trajetórias reais TID A A A A B B X 680271, 8508 680272, 0240 680271, 8575 680271, 5200 680271, 0138 680270, 0036 680269, 6661 680269, 6705 680269, 6772 Y 7462623, 6403 7462623, 8229 7462624, 1940 7462624, 5672 7462625, 1270 7462626, 4312 7462626, 8044 7462627, 1735 7462627, 7272 DATA 07 09 04 07 09 04 07 09 04 HORA 20 59 28 20 59 29 20 59 30 20 59 31 20 59 32 20 59 34 20 59 35 15 59 36 16 05 37 17

Trajectory Data q q q Spatio-temporal Data Represented by a set of points located in space and time (timestamped coordinates) T=(t 1, x 1, y 1), …, (tn, xn, yn) => position at time ti was (xi, yi) Fosca Giannotti 2007 – www. geopkdd. eu 18 18

Trajectories: Basic Concepts n Trajectories are represented by finite sequences of time-referenced locations, that result from various ways used to observe movements: • time-based recording: positions of entities are recorded at regularly spaced time moments, e. g. every 5 minutes; • change-based recording: a record is made when the position of an entity differs from the previous one; • location-based recording: records are made when an entity comes close to specific locations, e. g. where sensors are installed; • event-based recording: positions and times are recorded when certain events occur, in particular, activities performed by the moving entity (e. g. calling by a mobile phone); • various combinations of these basic approaches. Typically, positions are measured with uncertainty. Sometimes it is possible to refine the positions taking into account physical constraints, e. g. the street network. (Adrienko 2008) 19

Trajectories: Basic Concepts Movement-related characteristics include: • time, i. e. position of this moment on the time scale; • position of the entity in space; • direction of the entity’s movement; • speed of the movement (which is zero when the entity stays in the same place); • change of the direction (turn); • change of the speed (acceleration); • accumulated travel time and distance. (Adrienko 2008) 20 20

Trajectories: Overall Characteristics • geometric shape of the trajectory (fragment) in the space; • travelled distance, i. e. the length of the trajectory (fragment) in space; • duration of the trajectory (fragment) in time; • mean and maximal speed; • dynamics (behaviour) of the speed: – periods of constant speed, acceleration, deceleration; – characteristics of these periods: start and end times, duration, initial and final positions, initial and final speeds, etc. ; – arrangement (order) of these periods in time; (Adrienko 2008) 21 21

Trajectories: Overall Characteristics • dynamics (behaviour) of the directions: – periods of straight, curvilinear, circular movement; – characteristics of these periods: start and end times, initial and final positions and directions, major direction, angles of the curves, etc. ; – major turns (‘turning points’) with their characteristics: time, position, angle, initial and final directions, and speed of in the moment of the turn; (Adrienko 2008) 22 22

Relationships Generally, the goal of comparison is to establish relations between the objects that are compared. Here are some examples of possible relations: • equality or inequality; • order (less or greater, earlier or later, etc. ); • distance (in space, in time, or on any numeric scale); • topological relations (inclusion, overlapping, crossing, touching, etc. ). 23 23

Relationships Many other types of relations may be of interest, depending on the problem in hand: • similarity or difference of the overall characteristics of the trajectories (i. e. shapes, travelled distances, durations, dynamics of speed and directions, and so on); • spatial and temporal relations: – co-location in space, full or partial (i. e. the trajectories consist of the same positions or have some positions in common): · ordered co-location: the common positions are attained in the same order; · unordered co-location: the common positions are attained in different orders; – co-existence in time, full or partial (i. e. the trajectories are made during the same time period or the periods overlap); – co-incidence in space and time, full or partial (i. e. same positions are attained at the same time); – 24 24

Raw Trajectory Data: Problems and Solutions

The trajectory reconstruction problem n From raw location data (tid, x, y, t) a sample of a user’s movement (GPS recordings) n To trajectory data (obj-id, traj-id, (x, y, t)+) a sample of reconstructed trajectories (Theodoridis and Peleikis 2007) 26 26

Trajectory stream manager n Trajectory stream manager operations q q n receives raw location data about mobile users’ movement reconstructs trajectories (excluding noise, etc. ) and posts trajectory data to a MOD (Moving Object Database) Results so far – 2 alternatives q q Assumptions about trajectory ‘birth’ ( for spatial/temporal gaps between traces) Studying the notion of ‘stop’ (suspension of an entity’s movement) (Theodoridis and Peleikis 2007) 27 27

Trajectory stream manager (now…) (1) n When will an object have assigned a new trajectory-id? tn+1=10: 30 tm=11: 00 q q tn=10: 00 When there is sufficiently large gap in the spatial dimension between two consecutive recorded positions When there is sufficiently large gap in the temporal dimension between two consecutive recorded positions t 0=09: 00 t x (Theodoridis and Peleikis 2007) 28 28

Trajectory stream manager (now…) (2) n Dealing with noise q GPS-sampled positions may include noise, which should be excluded from trajectory reconstruction n A naïve approach computes the speed of the object in each segment of its motion and compares it with a commonly accepted maximum speed vmax (e. g. 200 km/h for cars) n In such a case, the stream manager rejects the last (marked as noisy) position and waits for the next (perhaps, acceptable) position to reconstruct a new segment (Theodoridis and Peleikis 2007) Vi>Vmax 29 29

BANCOS DE DADOS DE OBJETOS MOVEIS 30

O poder de BD de Objetos Móveis (Wolfson 1999) MOD n Passado: q Durante o ultimo ano, quantas vezes o ônibus 435 atrasou mais de 10 minutos ao passar pela parada 215? Restrição: Aeronaves devem voar a uma distância mínima de 2 km entre si. Futuro: Quais caminhões chegarão ao seu destino nos próximos 20 minutos? Presente: Onde estão os táxis a menos de 1 KM de onde estou? 31

Protótipos de Bancos de Dados de Objetos Móveis n SECONDO – Ralph Guting (Alemanha) n HERMES – Yannis Theodoridis and Nikos Pelekis (Grécia) 32

Secondo University of Hagen

Data Types (Guting 1999) Moving Point (mpoint) Moving Region (mregion) n n Data Types: mpoint e mregion são mapeamentos do tempo para o espaço q mpoint = ponto no tempo q mregion = região no tempo Exemplos: q vôo (id: string, origem: string, destino: string, rota: mpoint) q tempestade (id: string, tipo: string, area: mregion) 34

Operadores Espaço-Temporais (Guting 1999) Exemplos de Operadores: n Intersection (mpoint, mregion) → mpoint t 5 n distance (mpoint, mpoint) → mreal t 2 t 1 n trajectory (mpoint) → line n deftime(mpoint) → period t 0 n length (line) → real t 4 t 3 t 2 t 3 t 4 tn 35 t 5

Consultas Espaço-Temporais n vôo (id: string, origem: string, destino: string, rota: mpoint) n Consulta 1: “Encontre os vôos de São Paulo que voaram mais de 4000 km. ” SELECT * FROM voo WHERE origem = ’SP’ AND length (trajectory (rota) ) > 4000 n Consulta 2: “Encontre os pares de aviões que durante seus vôos se aproximaram em menos de 2000 metros!” SELECT f. id, g. id FROM voo f, voo g WHERE f. id <> g. id AND min (distance (f. rota, g. rota) ) < 2000 t 2 t 1 t 4 t 3 t 2 t 3 t 4 t 5 36

Hermes University of Pireaus

Hermes (Theodoridis and Peleikis 2007) n Dimensão espacial e temporal (tipo de dado PONTO) n HERMES Moving Data Cartridge (MDC) q n Implementado como um novo módulo, similar ao Oracle Spatial Data Cartridge Implementa diversos operadores espaco-temporais para relacionamentos espaço-temporais e similaridade: q Trajetórias individuais q Grupos de trajetórias 38

Operações n Gera um poligono ao redor um timestamp q n Calcula a distância entre dois pontos (tempo) de 2 objetos móveis q n f_front Verifica se um objeto está a atrás de um ponto em um certo instante de tempo q n f_distance Verifica se um objeto está a frente de um ponto em um certo instante de tempo q n f_buffer f_behind . . . rico grupo de operações espaciais 39

HERMES (Arquitetura) Temporal Dimension (Theodoridis and Peleikis 2007) Hermes Spatial Dimension 40

Projetos na Area de Trajetorias 41

Geo. PKDD – O PRIMEIRO projeto europeu na área de ANALISE trajetórias (2006 – 2009) Requisitos da Aplicação Visualização dos Dados e Padrões Teoria de BD Espaço-temporais Modelagem Conceitual e Ontologias Mineração de Dados Fornecimento Data Warehouse De e SGBD Dados Privacidade 42

MODAP– O SEGUNDO projeto europeu na área de trajetórias (2010 – 2012) 43

SEEK– PROJETO BRASIL – EUROPA (2012 – 2014) Universidades BRA: UFSC, PUC-Rio, UFSC, UFPE Universidades Europeias: Italia (Uni. Pisa, Uni. Veneza), Pireaus (Grecia) 44