3DeMoN ROBOVEC Integration of a new measuring instrument in an existing generic remote monitoring platform Luca Manetti 1 and Gilbert Steinmann 2 1 Smartec SA, via Pobbiette 11, 6928 Manno, Switzerland; manetti@smartec.ch 2 LMS-EPFL, Soil Mechanics Laboratory, EPFL Ecublens, 1015 Lausanne, Switzerland; gilbert.steinmann@epfl.ch ABSTRACT: The 3DeMoN (3-Dimentional Deformation Monitoring Network) monitoring system is a remote monitoring platform developed with the aim of offering a systematic and comprehensive approach to the solution of many environmental and civil engineering surveying and monitoring tasks. It integrates conventional measurement techniques and sensors (e.g. GPS receivers, Laser distance meters, Thermal Imaging Infrared Camera, etc.) with lowest power semiconductors, solar power supply, wireless communication, advanced database systems and Internet, with the aim of providing a unified management and view of remote monitoring data. This paper introduces the functional principles of this monitoring platform, presents its principal measuring technologies and illustrates the main characteristics of the latest developed and integrated instrument called ROBOVEC, a measuring unit based on a laser distance meter and a bi-axial modular robotization for horizontal and vertical movements. The first pilot installation on a landslide in Switzerland as well as the first monitoring results are presented. INTRODUCTION TO THE 3DEMON MONITORING PLATFORM The 3DeMoN system can be defined as an autonomous and automatic monitoring system consisting of a number of so called Monitoring Stations distributed on the object to be monitored (e.g. a landslide, a bridge or a building) and connected to one or more digital (e.g. GPS receivers or laser distance meters) or analogue sensors (e.g. inclinometers, load cells, etc.) and a so called Central Server Station (CSS). A 3DeMoN Monitoring Station consists of a weatherproof box containing the electronic components for the management of the sensors, the data transmission and the power supply modules. The Monitoring Stations communicate with the Central Server Station through a wired connection (RS-232/485, fiber optic) or though a wireless connection (GSM/GPRS/EDGE, radio modem). The Central Server Station Page 1
provides for data collection, data storage, management of the configuration of the Monitoring Stations, automatic data processing (e.g. automating batch postprocessing of GPS data) and monitoring of the correct operation of the network. Figure 1 shows the architecture of the 3DeMoN monitoring platform. Analogue sensor (e.g. piezometer) Digital sensor (e.g. GPS) Digital sensor (e.g. Laser) Software implemented: - Run Time Framework (RTF) - Open Management Framework (OMF) - Application Services 3DeMoN Monitoring Station Direct power Solar power Central Server Station (CSS) GSM / GPRS / EDGE or radio Wired connection Internet Front-end C ontrolc enter Software implemented: - Web Application - Database - General Services - e.g. GPS post-processing Software interfaces: - Web browser - Web application End-U ser Figure 1: 3DeMoN monitoring platform architecture Page 2
THE 3DEMON-GPS SYSTEM The first sensor that was integrated in the 3DeMoN platform was a single frequency L1 GPS receiver. This allows installing one or more autonomous GPS rover measuring stations on the object to be monitored plus one or more GPS reference stations on stable locations outside the monitored area. The carrier phase data are synchronously collected by all measuring stations and automatically uploaded on the Central Server Station. An automatic GPS static post-processing service performs all needed conversions, re-sampling and baseline calculations and returns the relative 3D coordinates of the monitored points as the final result of each measuring campaign. Figure 2: 3DeMoN-GPS rover measurement station with solar power supply Accuracy of the GPS system GPS satellites orbit around the earth with an orbiting period of 23 hours and 56 minutes. This means that approximately every 24 hours, over a location on the earth's surface, the satellites are disposed in the same spatial position (so called constellation) of the day before. As known, the spatial position of the satellites (DOP Dilution Of Precision) during the measurement time is one of the main factors influencing the accuracy of the results. Figure 3: Influence of the spatial satellite constellation (DOP) over 24h on the accuracy of GPS measurements (static post-processing solution) Page 3
Since the Central Server Station automatically shifts backwards the acquisition start time by 4 minutes every day, the DOP-induced error on the final coordinates can be considered as systematic and therefore neglected while repeating the same static measurements over the whole sensors network every 23h 56min. Typically the 3DeMoN-GPS system is programmed in order to perform 2-4 acquisition per day and the monitoring focuses on the daily averaged position of each point. The typical output of the system is shown in Figure 4, where the horizontal and vertical displacements of a point are monitored over a period of 12 months. Figure 4: Typical accuracy of the 3DeMoN-GPS system Raw data (blue), smoothed data (red) The performance of the system in summarized in the following table: Maximal baseline length 15 Km Typical acquisition time 20 min Accuracy of raw data: X, Y coordinate: ± 4mm Z coordinate: ± 9mm Accuracy of smoothed data: X, Y coordinate: ± 2mm Z coordinate: ± 4mm Table 1: Performance of the 3DeMoN-GPS system Page 4
THE 3DEMON-LASER SYSTEM The 3DeMoN-Laser is a laser distance meter based system that allows performing continuous and autonomous millimeter scale distance measurements for distances up to approximately 1,000 m. A 3DeMoN Monitoring Station manages up to 3 laser instruments through RS-232 connection providing power supply and automatic data transmission to the Central Server Station (CSS). The laser instruments are usually protected with a box (see Figure 5) in case of outdoors installation. When directly connected to the CSS (e.g. though an RS-422 bus) up to 10 Laser distance meters can be managed at the same time within a maximum distance of approximately 1 km. The 3DeMoN-Laser system can be used in several surveying applications such as land slides, earth settlement, slope instability, tunnel monitoring, tank level monitoring or civil engineering surveying. Figure 5: Installation with two 3DeMoN-Laser instruments (left) and a 3DeMoN Monitoring Station with solar power supply (right) Accuracy of the Laser system The typical accuracy of the laser instrument is ±1.5mm within maximum distances of about 800-1000m. The maximum measuring error on measurements relates to unfavorable conditions such as: high reflecting surfaces or heavy rough surfaces operating at the limits of the permitted temperature range or quick changes of ambient temperature very bright ambient conditions, strong heat shimmer excessive distance between instrument and measured point Indoor measurements Due to the stability of temperature and lighting conditions, indoor measurements are particularly accurate and stable. Accuracies up to 0.15 mm can be achieved within measured distances of about 40-50 m. Page 5
Figure 6: 3DeMoN-Laser, accuracy of indoor measurements Outdoors measurements The accuracy of outdoors measurements is affected by a combination of temperature, lighting conditions and other physical parameters of air such as humidity and pressure. Eventual additional visibility problems (mist, rain, frost, etc.) make the correction and interpretation of outdoors distance measurements slightly more difficult. The following figure shows the influence of the above-mentioned parameters on a measured distance of approximately 150m. Figure 7: 3DeMoN-Laser, accuracy of outdoor measurements Page 6
The performance of the system is summarized in the following table: Maximal measurable distance ~1,000 m Typical measurement time 0.2 4 s Typical accuracy (standard deviation) ±0.2 mm (indoor measurements) ±1.5 mm (outdoor measurements) Target needed up to 50m: none up to 200m: reflecting surface up to 1 km: geodetic prism Table 2: Performance of the 3DeMoN-Laser system THE 3DEMON CONTROL CENTRE SOFTWARE The 3DeMoN Control Center software is an integral part of the monitoring system and consists of all the software modules running on both Front-End and End-User levels (see Figure 1). Its development started to respond to the market s lack of flexible, scalable and robust software solutions to manage an extensive and heterogeneous network of measurement devices. In today s environmental and structural monitoring projects the information collected by conventional sensors, usually controlled by local data-loggers, is complemented or replaced by data acquired with more complex instruments like e.g. GPS receivers or Laser distancemeters. These devices are controlled by a more extensive set of parameters and the returned measurement data needs in some cases to be post-processed applying instrument specific algorithms or transformations. The 3DeMoN Control Center is a generic framework responsible for managing the configurations of the 3DeMoN remote Monitoring Stations and the single sensors or instruments connected to them, to establish manual or automatic connections to the stations, to remotely (re)configure the entire system, to download and locally store the measurement data, to trigger automatic post-processing and analysis activities on the collected data, to generate warnings and to publish results in different formats. The architecture of the 3DeMoN Control Center is based on a modern object- and component-oriented design and is implemented with the Java programming language, conforming to the Java 2 Enterprise Edition (J2EE) standards. The basic framework of the application can be extended, according to requirements of each single monitoring project, by pluggable software components specific for a certain communication device and protocol (e.g. GSM/GPSR/EDGE, Radio, RS-485 bus, etc.) or measurement instrument. Page 7
THE 3DEMON-ROBOVEC SYSTEM The ROBOVEC unit has been designed for (semi-)permanent installation in structural or geotechnical monitoring projects requiring a continuous monitoring of the 3-dimensional displacements of significant points under harsh environmental conditions. The ROBOVEC allows measuring distance, horizontal and vertical angles, temperature and reflected signal strength for an unidentified number of points. This Figure 8: ROBOVEC prototype information can be used to determine the variation of the positions of the monitored points. To date similar kind of monitoring programs are performed using conventional surveying instruments like e.g. high tech (and expensive) robot theodolites. The idea behind the development of the ROBOVEC was not to develop a copy of a robot theodolite, but to develop a measuring device fitting the needs of a certain number of monitoring applications where a simple laser distance meter does not fulfill the requirement and where performance and price of a robot theodolite are over-dimensioned. The following technical requirements were defined during the development phase of the ROBOVEC: robustness and possibility of outdoors operation without protection from atmospheric agents, (near)real-time data availability due to integration in existing 3DeMoN monitoring platform, possible integration of other sensors (e.g. a thermal Imaging Infrared Camera installed instead of the laser distance meter or a complementary Webcam, no need to be leveled ability to operate with any installation orientation, resolution of angle measurements: 0.001 deg, measurable distances: max. 1 km Figure 9: ROBOVEC prototype The ROBOVEC unit is based on a laser distance meter (right) and a bi-axial modular robotization (left) for horizontal and vertical orientation of the measuring device. The instrument is able to measure the variation of the distance and the variation of horizontal and vertical angles between the instrument zero and the monitored points. During a measurement program, the ROBOVEC unit can perform an auto-centering algorithm on each point in order to preserve the same targeting direction if the point has moved with respect to the previous measurement campaign. Page 8
Auto-centering algorithm While performing a full scan over the whole target's surface, the shape of the reflected signal strength behaves like a 3-dimensional Gaussian distribution. Applying different algorithms (e.g. a logarithmic regression analysis) on different horizontal and vertical scan profiles it is possible to determine the center of the target by calculating the respective maxima values. Figure 10 shows the 3D distribution of the signal strength reflected by a conventional reflector. Figure 10: 3D distribution of the signal strength reflected by a conventional reflector The following figure shows the typical shape of the reflected signal during a horizontal scan of a geodetic prism and the relative corrected logarithmic regression. Note the typical zero values in the middle of the target, consequence of a too high signal reflection. Figure 11: Typical profile of the reflected signal (purple) and relative corrected logarithmic regression (blue) Page 9
FIRST PILOT INSTALLATION The first ROBOVEC prototype was installed in June 2006 on the "La Frasse" landslide in the Vaud canton, Switzerland, not far from the city of Lausanne. A total of 11 points (9 on the landslide plus 2 reference points outside the sliding area) were equipped with a geodetic prism, as shown in Figure 14, and are measured every hour. This landslide is also instrumented with 4 inclinometers. These are manually measured every 2 months. Figure 12: View of the landslide from the ROBOVEC location As shown in Figure 13, the instrument was installed on a concrete pillar on the opposite side of the valley. The PC managing the ROBOVEC was equipped with a WiFi / GPRS / EDGE card, allowing a real-time remote access to the measurements. A direct connection to the 220 VAC power supply is available on-site. Figure 13: ROBOVEC instrument installed on pillar Figure 14: Geodetic prism on point Ref2 First results The distance measurements performed on the reference points (installed outside the moving area and therefore assumed to be stable) at the beginning of each measurement session are used to calculate the meteo-correction to be applied on the Page 10
measurements performed on the monitored points. This allows consideration and smoothing of the error induced by temperature, air pressure and humidity cited in the 3DeMoN-Laser system chapter. Figure 15: Displacements of the monitored points Figure 15 shows the results of a 4 month's period (October 2006 - February 2007). Point 1 and Point 8 present the highest sliding speeds with ~5 mm/month for Point 8 and ~4 mm/month for Point 1, starting from middle of November 2006. The averaged standard deviation of the measured distances is 1.2 mm. Figure 16: Targets and inclinometers positions Figure 16 shows the positions of the targets and the inclinometers. In particular P1 Page 11
and P8 are located on the most active area of the landslide and are therefore characterized by the largest displacements. Table below shows a comparison between the first results acquired with the ROBOVEC and the readings of the inclinometers. Particularly significant is the comparison between target and inclinometer P7-402 (orange) and P3-I301 (blue). ROBOVEC Targets Inclinometers P9 P6 P7 P5 P8 P1 P2 P3 P4 202 I301 402 LF406 2.2 5.5 1.8 7.7 19.9 15.7 9.0 5.0 3.0 5.0 6.5 0.5 6.0 Table 3: Displacements (in mm) from ROBOVEC and inclinometers readings CONCLUSIONS The most relevant concluding remarks about the development and the integration of the ROBOVEC instrument into the existing 3DeMoN monitoring platform are: The testing phase was successfully ended with the installation of the first pilot project on the "La Frasse" landslide in Switzerland The system is running automatically and independently from human intervention, providing in real-time through an Internet connection important data helping significantly the monitoring of this landslide The first monitoring results confirm that the 3DeMoN-ROBOVEC system is suitable for such an application, providing a reliable manner to track low dynamic displacements of natural objects such as landslides or earth settlements, or structures such as dams or bridges. REFERENCES Ingensand H. (1981). "Ein Beitrag zur Entwicklung eines elektronischen Tachymeters mit automatischer Zielerfassung", BDVI Forum. Steinmann G., Bonnard Ch. (2002). "Mesures en continu effectuées par le LMS sur des instabilités de terrains", Publication within CTI research project "FRAN2", Publication EPFL-LMS no S5919, 33 p. Bonnard Ch., Steinmann G. (1997), "A new laser distance meter for continuous measurement of landslide displacements", Rivista Italiana di Geotecnica, No 2/97, pp. 8-14 Tacher L., Bonnard Ch., Laloui L., Parriaux A. (2005). "Modelling the behaviour of a large landslide with respect to hydrogeological and geomechanical parameter heterogeneity". Journal of the International Consortium on Landslides, vol. 2, no 1, pp.3-14 Manetti L., Terribilini A., Knecht A. (2002). "Autonomous Remote Monitoring System for Landslides". SPIE's 9th annual international symposium on smart structures and materials, Proc. SPIE Vol. 4694, p. 230-235 Page 12