C C hih-Hong S hih-Hong S un, R un, R ong-K ong-K K K an an ang S ang S hang, J hang, J oe-Air J oe-Air J iang, iang, J ehn-Y ih J uang, , T zai-Hung Wen National T aiwa iwan University T T aiwan GI aiwan GI GIS GIS C C enter enter
Major C Major C hallenges hallenges to Our S to Our S ociet ociet iety iety S patially-enabled Intelligent C it ity MAK MAK OC OC I - Intelligent Decision C I - Intelligent Decision C C C loud loud P P rototype F rototype F ramework ramework Application C ase C C onclusions onclusions
E E nvironmental Problem nvironmental Problem E conomic Problem S S ocial Problem ocial Problem
1 exabyte = 1 billion gigabyte yte
A development based on information ion and communication technologies to improve decision making process to improve decision making process ss ss t t to solve problems to solve problems faced by our faced by our society http://www - 03.ibm.com/innovation/us/thesmartercity/ 03.ibm.com/innovation/us/thesmartercity/
dat a is geo-referenc ed geo-referenc ed
Use the concept of ambient in Use the concept of ambient in t intelligence t intelligence Integrate various Integrate various smart techno smart techno hnologies hnologies Include internet of things, geo eo-sensor network, wireless broadband, mobile device, an broadband, mobile device, an and cloud computing service, and cloud computing service, which have been embedded in d into our daily life
Cloud Com put in ing We We Web Service Web Service Data Data Go directly to the cloud and Go directly to the cloud and use web services use web services La Lack a platform to integrate data and mode an and transform them to information and an and transform them to information and kn knowledge for decision making
– P – P rovide robus rovide robus t s t s ervice by cloud-ins ervice by cloud-ins installation of data, models, and installation of data, models, and applications, allowing updating and and alternation among clouds. – Optimize the fit for decis – Optimize the fit for decis ion requir ion requir uirements uirements by integrating by integrating research research contribution from different institut tutes. – P – P roduce cus roduce cus tomized decis tomized decis ion s ion s upp upp pport s pport s ys ys tems tems according to user's according to user's requirements by utilizing multi-age gent and modularized data/models. C C reate a platform for reate a platform for “customizable” and “concurrently update ated” decision support systems decision support systems
An intelligent spatial decision An intelligent spatial decision ion support system cloud ion support system cloud computing service platform MAKOCI MAKOCI = M ulti- A gent K nowledge O r riented C yber I nfrastructure = Agent + Data + Knowledge = Agent + Data + Knowledge ge + Technology ge + Technology Intelligent Decision Cloud
Type Type S S ource ource Approach Approach Database Geospatial Data Gove overnment WMS , WF S , KML… S S ensor Data ensor Data Gove Gove overnment overnment S S WE WE Model Base Geospatial Indu dustry WP S , AP I F F unction unction Analysis Model R ese esearch institute Middleware
Ontology: “coded knowledge” Ontology: “coded knowledge” Organize knowledge and intelligen gence of human systematically. Build the relationship between da data and models. Agent: “coded reactions” Agent: “coded reactions” Quickly find out the information a n and analytical tools. T T rack changes, trigger the system, rack changes, trigger the system, m, and publish alerts. m, and publish alerts.
Data Cloud Data Cloud Model Cloud Model Cloud Data Data Model Analysis and Maps UI Design Decision Support System Workflow Workflow Simulation Outputs
Decision Model Geosp spatial Sensor Static support service functi ction data system system MAK MAK OC OC I can be used to: I can be used to: • R egister and organize data, mo model and web applications. • • F F acilitate developer to mash-u acilitate developer to mash-u up resources up resources into applications. into applications. • Help decision maker to look up up and use intelligent S DS S .
MAK MAK OC OC I can assist in understand I can assist in understand anding the change anding the change • The past: Historical data from m spatial database • The present: Monitoring data f ta from real-time sensor • The future: S imulation result fr lt from analysis model
Following the traditional way t Following the traditional way t y to develop y to develop a spatial decision support sys ystem System implemented get planning Sub- 3-5 years 5 years o Budgets Budgets Finish Finish contracts contracts developmen developmen developmen developmen Project Subcontracting Project Subcontracting (Including hardware, software, database, and user interface design) 30
Using the MAKOCI production Using the MAKOCI production tion platform to develop a tion platform to develop a spatial decision support syste stem uirements from Customized decision on Make contracts on leasing support system ecision makers Months of Months of equirements equirements Matching Matching Customizing Customizing Trial Trial Lease Lease development development development development The MAK OCI collaborating Test-r running the decision developers to match the developers to match the support system support system requirements 31
C C arry out a interdisciplinary proj arry out a interdisciplinary proj roject: roject: Including physical, human, engine ineering, computer, and GI S cience 1. E 1. E nabling the sensor web. nabling the sensor web. 2. Modeling for physical system. Applications 3. Modeling for human system. 3. Modeling for human system. 4. Integrating into a intelligent de ecision cloud and developing application systems developing application systems ms. ms. Sensor Data Sensor Data Physical Model Physical Model Human Model
Developing tiny smart sensor ne networks. Weather S ens Monitoring the temperature and temperature and humidity by wireless sensors
P P eople Number C eople Number C ounting ounting Detectingmovement by Detectingmovement by infrared light and delivering data by WiF i Access Point Database Internet User User
Modeling the spatial-temporal variat iations of land cover and urban micro-environment. micro-environment. E NVI-met E E NVI-met is NVI-met is a three-dimens a three-dimens ional microclim ional microclim limate model designed to limate model designed to simulate the surface-plant-air interactions ns in urban environment.
Present Expected It can be helpful for building design It can be helpful for building design and environmental improvement simulation. Temperature
T racking human contacts to simulate ate disease spread among high-dense populations populations by using geo-sensor web by using geo-sensor web eb and mobile sensors. eb and mobile sensors. Analyzing individual’s mobility and space-tim mobility and space-tim paths.
Movement d etecting Geo - sentinel sentinel node Real time Alarm Alarm indicators tiotemporal tiotemporal Modeling spatial Modeling spatial atabase diffusion of diseas Risk map of disease Risk map of disease transmission transmission Disease Disease transmission transmission transmission transmission transmission model
Building an intelligent spatial decision sion support platform which integrates integrates sensor data and modeling sensor data and modeling ing results. ing results.
ata: Quantitative precipitation forecast ata: Quantitative precipitation forecast Model: 2D Model: 2D inundation model inundation model
ule ata ata ion del Workflow: Linking data and mo model
Agent: S Agent: S etting rule and reactio etting rule and reactio ction ction 43
Decision support: E Decision support: E vacuation sugges vacuation sugges ulation result: Depth and time hydrograph ph
Create an opportunity for Create an opportunity for the industry of information services of an intelligent city, services of an intelligent city, collaborating companies, governments, and research Gove Gove vernment vernment Industry Industry R R esearch Instit esearch Instit governments, and research institutes. Publish geospatial Publish geospatial Develop decision Develop analytic and data sets data sets data sets data sets support systems support systems simulation models simulation models
Recommend
More recommend
