A Mobile Sensing Architecture for Massive Urban Scanning
0 likes969 views
The document proposes a mobile sensing architecture for massive urban scanning using sensors on public transportation vehicles and private vehicles to collect data. The data would be stored in a MySQL database and analyzed to increase safety and energy awareness. A mobile app would provide real-time feedback on fuel consumption. The system would benefit from cloud computing by enabling scalable data storage and analysis and providing an always-available web interface.
A Mobile Sensing Architecture for Massive Urban Scanning
- 1. A Mobile Sensing Architecture for Massive Urban Scanning João G. P. Rodrigues, Ana Aguiar, Fausto Vieira, João Barros, João P. Silva Cunha Massive Information Scavenging © 2010, IT - Instituto de Telecomunicações. Todos os direitos reservados.
- 2. Objectives and Motivation Massive Urban Scanning Hundreds of public transportation vehicles, can give us a constant and reliable stream of information Private vehicles complement the information with more roads and traffic awareness Increasing Safety and Energy Awareness on ITS By teaching drivers and companies after data analysis If you want to sense the environment… Roads are everywhere (that matters) Smartphones and sensors too! 2
- 3. Objectives 3
- 4. Architecture 4
- 5. Proposed System 5
- 6. Data Gathering: Sensors GPS -> Position WiFi -> Open spots, security analysis Bluetooth -> Population density Accelerometers -> Driving aggressiveness Video and audio -> Road events On-Board Diagnostics (OBD) Speed, RPM, Fuel Consumption, Accelerator pedal position Wearable Technologies – VitalJacket ECG -> HeartRate -> Stress 6
- 7. Data characteristics Basic data set of just 200 Bps 7
- 8. Database MySQL Relational DB Many joins GPS + … Many Writes Few reads 8
- 9. Initial Setup and Trials – Porto, Portugal 10 Probe Buses of the main Porto Carrier Data from GPS, Wifi, Bluetooth, Accelerometer and ECG 500 Taxis Real-time 1 Hz GPS traces and taxi status (On service, waiting, etc…) Volunteers with private cars Data from GPS, Wifi, Bluetooth, Accelerometer, OBD and ECG 9
- 10. Initial Setup and Trials – Porto, Portugal GPS 0.5s delayed, 0.2Km/h of average speed difference Before you drive into tunnels 10
- 11. Visualization Google Earth or Maps: KML file-based – Placemarks’ XML Very user friendly Spatial and Time correlations Many ways to show other data: Color of placemark Height Type 11
- 12. Fuel consumption analysis 12
- 13. Fuel Efficiency vs Speed 13
- 15. Mobile Application Collect GPS and Accelerometer data in real-time Feedback about fuel consumption and emissions in real-time OBD data may be used to calibrate model 15
- 16. Web Interface Be aware of own driving profile Calculate and compare consumptions for own trips Do I consume more or less than others in my vehicle category? Are there more efficient routes? 16
- 17. NIST Cloud Definition “Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.“ 17
- 18. Can you convince me that our project would profit from “cloud computing”? How? 18