P/00/011 Regulation 3.2 AUSTRALIA Patents Act 1990 ORIGINAL COMPLETE SPECIFICATION STANDARD PATENT Invention Title: METHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA Applicant: Intelematics Australia Pty Ltd The following statement is a full description of this invention, including the best method of performing it known to me: 1 6067 m 2 METHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA FIELD OF THE INVENTION 5 The present invention relates to the field of traffic information services. In particular the present invention relates to a method of normalising information such as Degree of Saturation (DoS) information from road sensors. The method may provide a relatively stable and less volatile data stream for use in traffic modelling and for producing traffic information. The method may include acquiring, processing and/or 10 producing traffic data. BACKGROUND OF THE INVENTION Traffic quality suffers when loop data from road sensors is volatile or is not stable. 15 Road sensor loop data may show interruption, hikes and general interruption in traffic flow which may require extensive and/or expensive correction of the data before it may be used in a traffic information service. The present invention may provide a dynamically adjustable traffic information service ?0 using normalised values of data within speed and travel time calculations to enhance quality and accuracy of available modelled road sensor data. Normalising unstable and volatile road sensor loop data may provide an easier data remediation path and may enhance and/or improve quality of a traffic service. The 25 road sensor loop data may include Degree of Saturation (DoS) data, being a measure of a ratio of effectively used green cycle time to total available green cycle time on an approach. SUMMARY OF THE INVENTION 30 The method of the present invention may provide a methodology that allows for adjustment of loop data from road sensors to correct volatility such as peaks and troughs in the loop data. The methodology may facilitate easy detection and/or correction of misrepresentations in the loop data as well as providing a more recent C \IoordMETHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA doc 3 trend or route profile. The methodology may also enhance and simplify coverage of a traffic service. The methodology of the present invention may use normalised DoS data values in 5 calculations of traffic congestion measures to provide more accurate information for users while utilising fewer resources. The normalised DoS data values may be integrated in real-time modelling of traffic information. The normalised values may provide a valuable opportunity for 10 enhancement and quality improvement in data calibration techniques. By dynamically normalising DoS data values for traffic model parameters using current real-time traffic conditions, a traffic information service may provide motorists with more accurate and tangible information on traffic conditions. It may also be more 15 responsive to irregular traffic trends on road segments, road sensor faults and propagation of gridlock in an associated road network. An improvement to quality of traffic information services may be achieved by integrating normalised DoS data values in a setting, validation and usage of traffic ?0 model parameters in real-time or in near real-time. The approach of the present invention may expand use of DoS data and its dynamic application to calibration of a traffic modelling process by applying techniques as described below. 25 An element of the process of the present invention includes a facility to provide statistically proven and quality enhanced calibration model parameters. According to the present invention there is provided a method for normalising 30 information from road sensors including raw Degree of Saturation (DoS) information, said method including: collecting said raw DoS information; constructing from said raw information a road DoS profile for each road sensor; constructing from said raw information a loop sites profile for each road sensor; and calculating a normalized CApof\word\ETHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATAcIC 4 DoS data set from said raw information, said road DoS profiles and said loop sites profiles. The step of constructing the road DoS profiles may include averaging DoS profiles for 5 roads with a similar lane configuration. The step of constructing the loop sites profiles may include grouping DoS values for sites with a similar arrangement of loop sensors on roads of a similar geometry. The step of calculating the normalized DoS data set may take into account current DoS information, road DoS profiles including minimum values, maximum values and range by type of day and time and loop site 10 characteristics including known faults, patterns and dependancies. The method may include a step of compensating for lack of data from defective road sensors. The method may use data from sensors at signalised or unsignalised intersections. The method may also include a step of compensating for differences in 15 raw DoS information collected from approaches with different numbers of loops. In a preferred embodiment data from sensor networks may be normalised by a methodology that includes at least the following steps: 1. collect raw DoS data from a coordinated adaptive traffic system such as SCATS 20 (Sydney Coordinated Adaptive Traffic System) via road loop sensors; 2. match geographically a respective road loop sensor site to a corresponding road link; 3. use knowledge about road type and configuration to adjust the raw DoS data; 4. use knowledge about a detector site to reduce noise from the raw DoS data; 25 5. group road segments and loop sensors by common characteristics; 6. use loop data profiles such as minimum/maximum values and variance of raw DoS data to differentiate between various loop data profiles over time; 7. apply factor analysis; 8. apply basic filtering to factor out compromised raw DoS data; 30 9. calculate predicted normalisation DoS values for the respective road loop sensors. Knowledge about road type and configuration may include topology information such as clearway information, strip shopping centres, school zones and tram lines. Noise C:\ow\METHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA.doC 5 from raw DoS data may be reduced by smoothing the data. In one form data may be smoothed by using a weighted output of the last 6 traffic cycle values. Road segments and loop sensors may be grouped based on similar road geometry 5 such as roads with a same number of lanes and/or layout, e.g. 2 lanes each way, 3 lanes each way, dual carriageway etc. Loop data profiles may be collected over time and mapped in order to give average, maximum and minimum DoS profiles. Real time loop profiles may then be compared 10 and normalised when a profile falls considerably out of range of the average. Factor analysis may include a statistical method used to describe variability among observed variables in terms of fewer unobserved variables known as factors. The observed variables may be modelled as linear combinations of the factors, plus "error" 15 terms. The information gained about interdependencies may be subsequently used to reduce the set of variables in a dataset. Basic filtering in one example may include removal of outliers such as samples falling below a 5 th percentile and beyond a 9 5 th percentile. Normalised DoS values may be ?0 calculated from profiles collected and mapped over time to provide average, maximum and minimum profiles for raw DoS data, road (processed) DoS data and loop site profiles. The raw values may be compared to historic values and normalised close to average for comparable days/times. 25 Loop site profiles may include raw and road (processed) DoS data values for sites with similar loop sensors arrangements on roads of similar geometry, i.e. roads with a similar number of detectors on approaches e.g. 2 detectors each way, 3 detectors each way, etc. 30 Through the above methodology, normalisation of DoS data values or indicators may provide a more stable and accurate source of real-time information. C \poword\WETHOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA.doc 6 DESCRIPTION OF A PREFERRED EMBODIMENT A preferred embodiment of the present invention will now be described with reference to the accompanying drawing wherein: 5 Figure 1 shows a flow diagram of a method for normalising DoS data obtained from road sensors. Referring to Figure 1, current traffic congestion indicators including raw DoS data (10) 10 is obtained from loop data (11) and is adjusted (12) using information about road type and configuration to construct DoS road profiles (13). Road configuration data (14) is used to geographically match loop sensors to road links while knowledge about the types of loops (15) is used to reduce noise from the raw DoS data. Road segments and loop sensors are grouped (16) by common characteristics to construct road loop 15 sites profiles (17). Loop data profiles including minimum/maximum values and variance of raw data, is applied to the road loop profiles (18). Basic filtering (19) is applied to the DoS road profiles (13) to factor out compromised data such as samples below a 5 th percentile 20 and beyond a 9 5 th percentile. Factor analysis (20) is applied to reduce variables in the road DoS profiles (13) and road loop profiles (17). A normalised DoS data set (21) is calculated from the raw DoS data (10), the DoS road profiles (13) and the road loop profiles (17) for each road 25 segment. The calculation takes into account current DoS readings, DoS profile characteristics including minimum, maximum values and range by type of day and time, and Loop Site characteristics including known faults, patterns and dependencies. 30 The loop data (11) is combined with other loop data (22) and the normalised DoS data (21) to produce traffic congestion data (23). The traffic congestion data (23) is subsequently processed and published (24) to provide enhanced traffic information. Caof.wrd\ME THOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA doc 7 Normalised DoS data may show more steady patterns than volatile raw DoS data. Using normalised DoS data in traffic modelling may minimise a need to apply aggressive smoothing of the traffic output. Normalised DoS data therefore saves time in processing of the data and hence saves costs. Use of normalised DoS data also 5 provides better traffic trend analysis and profile construction. Fields of application for the method of the present invention may include: - Traffic Analysis - Traffic Prediction 10 - Spatial Analysis and Integration with Traffic Information Services - GPS Route Calculation and Journey Statistics - Road Travel Time calculation - Road Speed Calculation - Device Monitoring and Tracking Data Manipulation 15 - Database Systems. Finally, it is to be understood that various alterations, modifications and/or additions may be introduced into the constructions and arrangements of parts previously described without departing from the spirit or ambit of the invention. C apofwrd\ME THOD FOR NORMALISING INFORMATION FROM TRAFFIC DATA doc