SlideShare uma empresa Scribd logo
1 de 35
Baixar para ler offline
X-Sense
Monitoring Alpine Mass Movements at Multiple Scales
- Annual Meeting 13   th   May 2011   -

Lothar Thiele, Jan Beutel                 ETH Zurich, Embedded/Wireless
Stephan Gruber                            University Zurich, Physical Geography
Alain Geiger                              ETH Zurich, Geodesy and Photogrammetry
Tazio Strozzi, Urs Wegmüller              GAMMA SA, SAR Remote Sensing
Hugo Raetzo                               BAFU/FOEN

                                              1
[Eiger east-face rockfall, July 2006, images courtesy of Arte Television]


                                                                     2
X-Sense Hypothesis

 Anticipation of future environmental states and risk
   is improved by
     a systematic combination of environmental sensing at
       diverse temporal and spatial scales and
     process modeling

 Wireless Sensor Network Technology
     allows to quantify mountain cryosphere phenomena and their
      transient response to climate change
     can be used for safety critical applications in an hostile
      environment


                                3
Climate change and cryosphere as
(additional) elements of surprise




                   4
New Avenues for X-Sense
   Detecting and measuring large-scale terrain movement
    Understanding newly-developed slope movements

   Current methods:                                      > 100 cm/year
    InSAR measurements                                  50-100 cm/year
                                                         10-50 cm/year
    Manual D-GPS                                        2-10 cm/year
                                                         0-2 cm/year



                                   Sensor challenges
                                    Complex sensors (combinations
                                     of sensors, different scales)
                                    Variable data rates
                                    User interaction (feedback)
                                    In-network processing
                               5
X-Sense Platform
                                              Host Station
                                              processing, fusion, storage




      Reference GPS




                      Moving debris
                      moving rock slope
                                          6
Sensor Network Promises
 Sensor nodes are cheap, so we can have plenty of them.
 Nodes may be cheap, but deployment and maintenance is
 expensive.



 Additional redundant nodes make the system fault tolerant
 automatically.
 More nodes make the system more fragile.



 End-to-end Predictability and Efficiency
                             7
- Design Approach –


Develop a methodology for the design of
 dependable wireless sensor networks




                    8
Challenge: The Physical Environment
   Lightning, avalanches, rime,
   prolonged snow/ice cover, rockfall
   Strong daily variation of temperature
    −30 to +40°C
    ∆T ≦ 20°C/hour




                               9
Challenge: The Design Approach
   Traditional iterative design approach: waterfall-model
   Repeated for individual system layers




  Testbed         [Matthias Woehrle]


  insufficient knowledge of target application / environment
  working on resource limits
                                       10
Top-down Approach: In-situ Design & Test
    Feature-rich Platform
                                                     Refined
                              Behavioral Data        Platform
                                                   Specification
                            observe,
                            experiment,
                            learn on-site


  Flexible in-situ exploration (testbed ≠ real system)
  Real sensor data, real environment
  Integration with live data management (system of systems)
                               11
- Deployment –

 Provide a prototype system that allows to
 quantify mountain cryosphere phenomena
and can be used in early warning scenarios.




                      12
Field Site Selection




                       13
Vanessa Wirz
                                                           Vanessa Wirz

Location Planning of Measurement Devices
•TerraSAR-X
                                               Field site selection based
•(Sept. 2009, 11 days)                         on aerial photographs,
                                               satellite-based InSAR
                                               detection and fieldwork




                     •reference devices




     •Dirru rock glacier
     •velocity > 1 m/a




                                          14
Bernhard Buchli
                                           Tonio Gsell, Christoph Walser
New GPS Logger Devices                     Roman Lim, Mustafa Yucuel

   30 GPS logger devices have been
   designed and manufactured in
   partnership with Art-of-Technology AG
   Financially supported by BAFU/FOEN
   and canton Wallis
   Deployment started Q4/2010




                             15
Current Test Deployment in Valais




                    16
Wireless Infrastructure Randa/Dirruhorn
   20 km WLAN link from Zermatt to Randa
    Collaboration with CCES projects: APUNCH + COGEAR (P.
     Burlando; ETHZ, S. Loew)
   Longest low-power wireless sensor network link
    Uses TinyNode184 and directional antenna
    Stable operation since 08/2010




                              17
Current and Planned Installation




                    18
- Methodology –

Provide methods and tools for the design of a
dependable, long-term sensing infrastructure
         in extreme environments.




                       19
Ultra Low-Power Multi-hop Networking
   Dozer ultra low-power data gathering system                          [Burri, IPSN2007]
     Beacon based, 1-hop synchronized TDMA
     Optimized for ultra-low duty cycles
     0.167% duty-cycle, 0.032mA (@ 30sec beacons)
           contention
            window         data transfer                     beacon

                                                    jitter

               slot 1   slot 2             slot k                time


   But in reality: Connectivity can not be guaranteed…
    Situation dependent transient links (scans/re-connects use energy)
    Account for long-term loss of connectivity (snow!)




                                                    20
Challenge: Low Power Operation




                  21
Formal Conformance Test   Matthias Woehrle




                  22
Formal Conformance Test

            •Model of                      •Model of
                           •Verify
            observed    Reachability in    expected
            behavior                       behavior
•Power                    UPPAAL
 trace        •PT                            •Sys




 •System in operation          •Expected behavior


                                              •[FORMATS 2009]

                        23
Challenge: Data Integrity                  Matthias Keller

 •   Long term deployment
 •   Up to 19 sensor nodes
 •   TinyOS/Dozer [Burri, IPSN2007]
 •   Constant rate sampling
 •   < 0.1 MByte/node/day




                                      24
Data is not Correct-by-Design
     Artifacts observed
        Packet duplicates
        Packet loss
        Wrong ordering
        Variations in received vs. expected packet rates
     Necessitates further data cleaning/validation




                                  25
Sources of Errors included in Model
Data Loss                                                             ^
                                                 Clock Drift ρ [ -ρ; +ρ]  ^

               Node reboot                       Directly affects measurement of
                                                      • Sampling period T
                  ✗                                   • Contribution to elapsed time te
                   ✗✗
                                                 Indirectly leading to inconsistencies
Waiting        Queue reset          Empty
packets                             queue
                                                      • Time stamp order tp vs. order of
                                                        packet generation s

Packet Duplicates                                Node Restarts
                                                 • Cold restart: Power cycle
                 2 Lost 1-hop ACK
             ✗                                   • Warm restart: Watchdog reset
                     1

                                                            T          <T
                 3                               • Shortens packet period
          Retransmission                         • Resets/rolls over certain counters

                                            26
Model-based Data Validation Case Study
   Reconstruction
   of correct temporal
   order

   Validation of correct
   system function

   Domain user
   interested
   in “correct” data


                                 [Keller, IPSN2011]
                           27
- Data Processing –

Develop models and algorithms that process
   multi-scale data and allow to quantify
    mountain cryosphere phenomena.




                     28
GPS Data Analysis
  Challenges
    Processing strategies
    Optimal duty-cycle strategy
    Near real-time GPS
     processing techniques

  Continuous observations of surface motion with low cost GPS
    Differential L1 carrier phase post-processing and velocity estimation
     based on piecewise polynomial fit.
    Reliable observation of velocities < 2 cm/day

Continuous GPS monitoring reveals velocity changes at high
temporal resolution strongly correlated with ambient parameters.

                                     29
GPS Testbed
                                    •Kinematic positioning error [m]
 •GPS positions (unfiltered)




  •Velocity




                 •15 months         [Limpach, GGL, 2011]
                               30
Measured displacement rate and
 simulated ground temperature




                            Stefano Endrizzi
              31
Measured displacement rate and
 simulated soil water content




                            Stefano Endrizzi
              32
Data Fusion of GPS and InSAR
   Idea
    Quasi continuous observations of surface motion with low cost GPS
    SAR satellite measurements cover surface area at certain time
     epochs (SAR data processing by GAMMA)
    Data fusion between continuous GPS velocity field at receiver
     locations and InSAR displacement field in LOS between specific
     time epochs
   Ongoing Developments
     Modeling 3-D surface displacement field based on GPS results
     Incorporate 1-D InSAR displacement field
     Increase model accuracy using different filter techniques
     Development of time dependent surface movement using accurate
      DTM
     Computation of strain and stress fields

                                 33
Data Fusion of GPS and InSAR

High resolution GPS
stations provide a
quasi continuous
observation of surface
points.

SAR images can be
used to extend and
improve the surface
motion modelling in
the area of interest at
any point in time.


                               [Neyer, GGL, 2011]
                          34
•     ETH Zurich
       –   Computer Engineering and Networks Lab
       –   Geodesy and Geodynamics Lab
•     University of Zurich
       –   Department of Geography
•     Gamma SA
       –   SAR Remote Sensing
•     BAFU/FOEN
       –   Federal Office for the Environment




Interested in more?
http://www.permasense.ch
 35

Mais conteúdo relacionado

Mais procurados

DRX1 Brochure
DRX1 BrochureDRX1 Brochure
DRX1 Brochurel126073
 
An image based disdrometer verification and raindrop analysis
An image based disdrometer verification and raindrop analysisAn image based disdrometer verification and raindrop analysis
An image based disdrometer verification and raindrop analysisJames Huang
 
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...High Performance Cyberinfrastructure Required for Data Intensive Scientific R...
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...Larry Smarr
 
20th. Single Molecule Workshop Picoquant 2014
20th. Single Molecule Workshop Picoquant 201420th. Single Molecule Workshop Picoquant 2014
20th. Single Molecule Workshop Picoquant 2014Dirk Hähnel
 
Thermal network cameras Performance considerations for intelligent video
Thermal network cameras Performance considerations for intelligent videoThermal network cameras Performance considerations for intelligent video
Thermal network cameras Performance considerations for intelligent videoAxis Communications
 

Mais procurados (8)

DRX1 Brochure
DRX1 BrochureDRX1 Brochure
DRX1 Brochure
 
PAC-Grenoble: Radiation hardness testing, Industry Case Study: Airbus
PAC-Grenoble: Radiation  hardness testing, Industry Case Study: AirbusPAC-Grenoble: Radiation  hardness testing, Industry Case Study: Airbus
PAC-Grenoble: Radiation hardness testing, Industry Case Study: Airbus
 
An image based disdrometer verification and raindrop analysis
An image based disdrometer verification and raindrop analysisAn image based disdrometer verification and raindrop analysis
An image based disdrometer verification and raindrop analysis
 
Als seminar
Als seminarAls seminar
Als seminar
 
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...High Performance Cyberinfrastructure Required for Data Intensive Scientific R...
High Performance Cyberinfrastructure Required for Data Intensive Scientific R...
 
SECURE: Semantics Empowered resCUe enviRonmEnt
SECURE: Semantics Empowered resCUe enviRonmEntSECURE: Semantics Empowered resCUe enviRonmEnt
SECURE: Semantics Empowered resCUe enviRonmEnt
 
20th. Single Molecule Workshop Picoquant 2014
20th. Single Molecule Workshop Picoquant 201420th. Single Molecule Workshop Picoquant 2014
20th. Single Molecule Workshop Picoquant 2014
 
Thermal network cameras Performance considerations for intelligent video
Thermal network cameras Performance considerations for intelligent videoThermal network cameras Performance considerations for intelligent video
Thermal network cameras Performance considerations for intelligent video
 

Semelhante a Xsense

Towards the Wikipedia of World Wide Sensors
Towards the Wikipedia of World Wide SensorsTowards the Wikipedia of World Wide Sensors
Towards the Wikipedia of World Wide SensorsCybera Inc.
 
Semantically enabled standard development
Semantically enabled standard developmentSemantically enabled standard development
Semantically enabled standard developmentLaurent Lefort
 
A benchmark dataset to evaluate sensor displacement in activity recognition
A benchmark dataset to evaluate sensor displacement in activity recognitionA benchmark dataset to evaluate sensor displacement in activity recognition
A benchmark dataset to evaluate sensor displacement in activity recognitionOresti Banos
 
Semantics in Sensor Networks
Semantics in Sensor NetworksSemantics in Sensor Networks
Semantics in Sensor NetworksOscar Corcho
 
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game Changer
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game ChangerAdvanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game Changer
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game ChangerSabrie Soloman
 
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...Jason Edwards
 
Wireless Sensor Networks UNIT-1
Wireless Sensor Networks UNIT-1Wireless Sensor Networks UNIT-1
Wireless Sensor Networks UNIT-1Easy n Inspire L
 
Hsis2005 Geospatial Nomadeyes Full
Hsis2005 Geospatial Nomadeyes FullHsis2005 Geospatial Nomadeyes Full
Hsis2005 Geospatial Nomadeyes Fullmartindudziak
 
Sensor Networks Introduction and Architecture
Sensor Networks Introduction and ArchitectureSensor Networks Introduction and Architecture
Sensor Networks Introduction and ArchitecturePeriyanayagiS
 
Drones and A.I in Earth Science
Drones and A.I in Earth ScienceDrones and A.I in Earth Science
Drones and A.I in Earth ScienceARDC
 
Mid Term Demo
Mid Term DemoMid Term Demo
Mid Term Demoodcsss08
 
C13_GaugeKeeper_englisch
C13_GaugeKeeper_englischC13_GaugeKeeper_englisch
C13_GaugeKeeper_englischEbi Jose, PhD
 
Hybrid Target Tracking Scheme in Wireless Sensor Networks
Hybrid Target Tracking Scheme in Wireless Sensor NetworksHybrid Target Tracking Scheme in Wireless Sensor Networks
Hybrid Target Tracking Scheme in Wireless Sensor NetworksIRJET Journal
 
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networks
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor NetworksWiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networks
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networkssmartcameras
 

Semelhante a Xsense (20)

Darema - Dynamic Data Driven Applications Systems (DDDAS) - Spring Review 2013
Darema - Dynamic Data Driven Applications Systems (DDDAS) - Spring Review 2013Darema - Dynamic Data Driven Applications Systems (DDDAS) - Spring Review 2013
Darema - Dynamic Data Driven Applications Systems (DDDAS) - Spring Review 2013
 
Senslab - open hardware - fossa2010
Senslab - open hardware - fossa2010Senslab - open hardware - fossa2010
Senslab - open hardware - fossa2010
 
Towards the Wikipedia of World Wide Sensors
Towards the Wikipedia of World Wide SensorsTowards the Wikipedia of World Wide Sensors
Towards the Wikipedia of World Wide Sensors
 
Unit 2-basic wireless sensor
Unit 2-basic wireless sensorUnit 2-basic wireless sensor
Unit 2-basic wireless sensor
 
Semantically enabled standard development
Semantically enabled standard developmentSemantically enabled standard development
Semantically enabled standard development
 
842 manobianco
842 manobianco842 manobianco
842 manobianco
 
A benchmark dataset to evaluate sensor displacement in activity recognition
A benchmark dataset to evaluate sensor displacement in activity recognitionA benchmark dataset to evaluate sensor displacement in activity recognition
A benchmark dataset to evaluate sensor displacement in activity recognition
 
Semantics in Sensor Networks
Semantics in Sensor NetworksSemantics in Sensor Networks
Semantics in Sensor Networks
 
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game Changer
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game ChangerAdvanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game Changer
Advanced Sensor Tech : SpectRx™ , Lightning Speed & Hardness Test Game Changer
 
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...
10121-4504-01-PR-Final_Intelligent_Casing-Intelligent_Formation_Telemetry_Sys...
 
Wireless Sensor Networks UNIT-1
Wireless Sensor Networks UNIT-1Wireless Sensor Networks UNIT-1
Wireless Sensor Networks UNIT-1
 
Hsis2005 Geospatial Nomadeyes Full
Hsis2005 Geospatial Nomadeyes FullHsis2005 Geospatial Nomadeyes Full
Hsis2005 Geospatial Nomadeyes Full
 
Sensor Networks Introduction and Architecture
Sensor Networks Introduction and ArchitectureSensor Networks Introduction and Architecture
Sensor Networks Introduction and Architecture
 
Drones and A.I in Earth Science
Drones and A.I in Earth ScienceDrones and A.I in Earth Science
Drones and A.I in Earth Science
 
Mid Term Demo
Mid Term DemoMid Term Demo
Mid Term Demo
 
Sensor net
Sensor netSensor net
Sensor net
 
C13_GaugeKeeper_englisch
C13_GaugeKeeper_englischC13_GaugeKeeper_englisch
C13_GaugeKeeper_englisch
 
Hybrid Target Tracking Scheme in Wireless Sensor Networks
Hybrid Target Tracking Scheme in Wireless Sensor NetworksHybrid Target Tracking Scheme in Wireless Sensor Networks
Hybrid Target Tracking Scheme in Wireless Sensor Networks
 
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networks
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor NetworksWiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networks
WiSE-MNet: an experimental environment for Wireless Multimedia Sensor Networks
 
Sensor.ppt
Sensor.pptSensor.ppt
Sensor.ppt
 

Mais de dalgetty

Mais de dalgetty (20)

Analyse fr
Analyse frAnalyse fr
Analyse fr
 
Analyse eng
Analyse engAnalyse eng
Analyse eng
 
Cmosaic
CmosaicCmosaic
Cmosaic
 
NanowireSensor (Nano-Tera)
NanowireSensor (Nano-Tera)NanowireSensor (Nano-Tera)
NanowireSensor (Nano-Tera)
 
Nexray
NexrayNexray
Nexray
 
SImOS
SImOSSImOS
SImOS
 
Irsens
Irsens Irsens
Irsens
 
cmosaic
cmosaiccmosaic
cmosaic
 
NanowireSensor
NanowireSensorNanowireSensor
NanowireSensor
 
LiveSense
LiveSenseLiveSense
LiveSense
 
Mixsel
MixselMixsel
Mixsel
 
Cabtures
CabturesCabtures
Cabtures
 
Greenpower
GreenpowerGreenpower
Greenpower
 
Placitus
PlacitusPlacitus
Placitus
 
Patlisci
PatlisciPatlisci
Patlisci
 
QCrypt
QCryptQCrypt
QCrypt
 
Nutrichip
NutrichipNutrichip
Nutrichip
 
I-ironic
I-ironicI-ironic
I-ironic
 
Selfsys
SelfsysSelfsys
Selfsys
 
ISyPeM
ISyPeMISyPeM
ISyPeM
 

Último

"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii SoldatenkoFwdays
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLScyllaDB
 
DSPy a system for AI to Write Prompts and Do Fine Tuning
DSPy a system for AI to Write Prompts and Do Fine TuningDSPy a system for AI to Write Prompts and Do Fine Tuning
DSPy a system for AI to Write Prompts and Do Fine TuningLars Bell
 
Advanced Computer Architecture – An Introduction
Advanced Computer Architecture – An IntroductionAdvanced Computer Architecture – An Introduction
Advanced Computer Architecture – An IntroductionDilum Bandara
 
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxMerck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxLoriGlavin3
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024Lorenzo Miniero
 
Scanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL CertsScanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL CertsRizwan Syed
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Enterprise Knowledge
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxhariprasad279825
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfAlex Barbosa Coqueiro
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupFlorian Wilhelm
 
The Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsThe Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsPixlogix Infotech
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfAddepto
 
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Commit University
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyAlfredo García Lavilla
 
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Mark Simos
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr BaganFwdays
 

Último (20)

"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQL
 
DSPy a system for AI to Write Prompts and Do Fine Tuning
DSPy a system for AI to Write Prompts and Do Fine TuningDSPy a system for AI to Write Prompts and Do Fine Tuning
DSPy a system for AI to Write Prompts and Do Fine Tuning
 
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptxE-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
 
Advanced Computer Architecture – An Introduction
Advanced Computer Architecture – An IntroductionAdvanced Computer Architecture – An Introduction
Advanced Computer Architecture – An Introduction
 
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxMerck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024
 
Scanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL CertsScanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL Certs
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
 
The Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsThe Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and Cons
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
 
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easy
 
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special EditionDMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
 
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan
 

Xsense

  • 1. X-Sense Monitoring Alpine Mass Movements at Multiple Scales - Annual Meeting 13 th May 2011 - Lothar Thiele, Jan Beutel ETH Zurich, Embedded/Wireless Stephan Gruber University Zurich, Physical Geography Alain Geiger ETH Zurich, Geodesy and Photogrammetry Tazio Strozzi, Urs Wegmüller GAMMA SA, SAR Remote Sensing Hugo Raetzo BAFU/FOEN 1
  • 2. [Eiger east-face rockfall, July 2006, images courtesy of Arte Television] 2
  • 3. X-Sense Hypothesis Anticipation of future environmental states and risk is improved by  a systematic combination of environmental sensing at diverse temporal and spatial scales and  process modeling Wireless Sensor Network Technology  allows to quantify mountain cryosphere phenomena and their transient response to climate change  can be used for safety critical applications in an hostile environment 3
  • 4. Climate change and cryosphere as (additional) elements of surprise 4
  • 5. New Avenues for X-Sense Detecting and measuring large-scale terrain movement  Understanding newly-developed slope movements Current methods: > 100 cm/year  InSAR measurements 50-100 cm/year 10-50 cm/year  Manual D-GPS 2-10 cm/year 0-2 cm/year Sensor challenges  Complex sensors (combinations of sensors, different scales)  Variable data rates  User interaction (feedback)  In-network processing 5
  • 6. X-Sense Platform Host Station processing, fusion, storage Reference GPS Moving debris moving rock slope 6
  • 7. Sensor Network Promises Sensor nodes are cheap, so we can have plenty of them. Nodes may be cheap, but deployment and maintenance is expensive. Additional redundant nodes make the system fault tolerant automatically. More nodes make the system more fragile. End-to-end Predictability and Efficiency 7
  • 8. - Design Approach – Develop a methodology for the design of dependable wireless sensor networks 8
  • 9. Challenge: The Physical Environment Lightning, avalanches, rime, prolonged snow/ice cover, rockfall Strong daily variation of temperature  −30 to +40°C  ∆T ≦ 20°C/hour 9
  • 10. Challenge: The Design Approach Traditional iterative design approach: waterfall-model Repeated for individual system layers Testbed [Matthias Woehrle]  insufficient knowledge of target application / environment  working on resource limits 10
  • 11. Top-down Approach: In-situ Design & Test Feature-rich Platform Refined Behavioral Data Platform Specification observe, experiment, learn on-site Flexible in-situ exploration (testbed ≠ real system) Real sensor data, real environment Integration with live data management (system of systems) 11
  • 12. - Deployment – Provide a prototype system that allows to quantify mountain cryosphere phenomena and can be used in early warning scenarios. 12
  • 14. Vanessa Wirz Vanessa Wirz Location Planning of Measurement Devices •TerraSAR-X Field site selection based •(Sept. 2009, 11 days) on aerial photographs, satellite-based InSAR detection and fieldwork •reference devices •Dirru rock glacier •velocity > 1 m/a 14
  • 15. Bernhard Buchli Tonio Gsell, Christoph Walser New GPS Logger Devices Roman Lim, Mustafa Yucuel 30 GPS logger devices have been designed and manufactured in partnership with Art-of-Technology AG Financially supported by BAFU/FOEN and canton Wallis Deployment started Q4/2010 15
  • 16. Current Test Deployment in Valais 16
  • 17. Wireless Infrastructure Randa/Dirruhorn 20 km WLAN link from Zermatt to Randa  Collaboration with CCES projects: APUNCH + COGEAR (P. Burlando; ETHZ, S. Loew) Longest low-power wireless sensor network link  Uses TinyNode184 and directional antenna  Stable operation since 08/2010 17
  • 18. Current and Planned Installation 18
  • 19. - Methodology – Provide methods and tools for the design of a dependable, long-term sensing infrastructure in extreme environments. 19
  • 20. Ultra Low-Power Multi-hop Networking Dozer ultra low-power data gathering system [Burri, IPSN2007]  Beacon based, 1-hop synchronized TDMA  Optimized for ultra-low duty cycles  0.167% duty-cycle, 0.032mA (@ 30sec beacons) contention window data transfer beacon jitter slot 1 slot 2 slot k time But in reality: Connectivity can not be guaranteed…  Situation dependent transient links (scans/re-connects use energy)  Account for long-term loss of connectivity (snow!) 20
  • 21. Challenge: Low Power Operation 21
  • 22. Formal Conformance Test Matthias Woehrle 22
  • 23. Formal Conformance Test •Model of •Model of •Verify observed Reachability in expected behavior behavior •Power UPPAAL trace •PT •Sys •System in operation •Expected behavior •[FORMATS 2009] 23
  • 24. Challenge: Data Integrity Matthias Keller • Long term deployment • Up to 19 sensor nodes • TinyOS/Dozer [Burri, IPSN2007] • Constant rate sampling • < 0.1 MByte/node/day 24
  • 25. Data is not Correct-by-Design Artifacts observed  Packet duplicates  Packet loss  Wrong ordering  Variations in received vs. expected packet rates Necessitates further data cleaning/validation 25
  • 26. Sources of Errors included in Model Data Loss ^ Clock Drift ρ [ -ρ; +ρ] ^ Node reboot Directly affects measurement of • Sampling period T ✗ • Contribution to elapsed time te ✗✗ Indirectly leading to inconsistencies Waiting Queue reset Empty packets queue • Time stamp order tp vs. order of packet generation s Packet Duplicates Node Restarts • Cold restart: Power cycle 2 Lost 1-hop ACK ✗ • Warm restart: Watchdog reset 1 T <T 3 • Shortens packet period Retransmission • Resets/rolls over certain counters 26
  • 27. Model-based Data Validation Case Study Reconstruction of correct temporal order Validation of correct system function Domain user interested in “correct” data [Keller, IPSN2011] 27
  • 28. - Data Processing – Develop models and algorithms that process multi-scale data and allow to quantify mountain cryosphere phenomena. 28
  • 29. GPS Data Analysis Challenges  Processing strategies  Optimal duty-cycle strategy  Near real-time GPS processing techniques Continuous observations of surface motion with low cost GPS  Differential L1 carrier phase post-processing and velocity estimation based on piecewise polynomial fit.  Reliable observation of velocities < 2 cm/day Continuous GPS monitoring reveals velocity changes at high temporal resolution strongly correlated with ambient parameters. 29
  • 30. GPS Testbed •Kinematic positioning error [m] •GPS positions (unfiltered) •Velocity •15 months [Limpach, GGL, 2011] 30
  • 31. Measured displacement rate and simulated ground temperature Stefano Endrizzi 31
  • 32. Measured displacement rate and simulated soil water content Stefano Endrizzi 32
  • 33. Data Fusion of GPS and InSAR Idea  Quasi continuous observations of surface motion with low cost GPS  SAR satellite measurements cover surface area at certain time epochs (SAR data processing by GAMMA)  Data fusion between continuous GPS velocity field at receiver locations and InSAR displacement field in LOS between specific time epochs Ongoing Developments  Modeling 3-D surface displacement field based on GPS results  Incorporate 1-D InSAR displacement field  Increase model accuracy using different filter techniques  Development of time dependent surface movement using accurate DTM  Computation of strain and stress fields 33
  • 34. Data Fusion of GPS and InSAR High resolution GPS stations provide a quasi continuous observation of surface points. SAR images can be used to extend and improve the surface motion modelling in the area of interest at any point in time. [Neyer, GGL, 2011] 34
  • 35. ETH Zurich – Computer Engineering and Networks Lab – Geodesy and Geodynamics Lab • University of Zurich – Department of Geography • Gamma SA – SAR Remote Sensing • BAFU/FOEN – Federal Office for the Environment Interested in more? http://www.permasense.ch 35