Analysis of Bruntingthorpe crash test data

Transcription

1 Analysis of Bruntingthorpe crash test data Martin Worrell Technical Director : APD Introduction APD has provided mission-critical and businesscritical solutions to organisations within the public sector, transport, security and logistics industries for over 30 years. APD's products are used in over 100 client sites in the United Kingdom, Scandinavia, Eastern Europe and the Middle East. Its control room, mobile information, resource location and tracking solutions are in daily operational use in every UK Police Force. With offices in the UK and the Middle East, APD is a global leader in the supply of proven and robust communications-based technology. In addition to the solid presence in UK Emergency Services organisations, other customers include the Swedish Police, Abu Dhabi Police and leading organisations such as The Emirates Group and London Underground Limited. APD s three major product lines are : - ARTEMIS : Vehicle Tracking & Telematics CORTEX : Integrated Communications Control System (ICCS) ASPIRE : Public Contact Management Brief Overview of ARTEMIS APD s ARTEMIS solution provides a broad range of cross-functional benefits to many departments within a Police Force. performance becomes a cause for concern. This enables speedy and targeted intervention and has been proven to reduce the number of vehicle collisions a Force experiences. Collisions inevitably, still occur, and when they do ARTEMIS provides all of the necessary functionality to deal with them quickly and efficiently. ARTEMIS therefore provides a comprehensive Incident Data Recorder (IDR) and in this article we will be focussing on the IDR module and how it was used to capture and analyse data from Crash Day Superseding the robust and proven platform of its predecessor, INCA CoOrdinator, ARTEMIS was developed to meet the increasing demands of the UK Police, in line with the ACPO ITS One Box Project. INCA was APD s flagship in-vehicle blackbox solution, designed in , and supplied to Police Forces between 1998 and Many of you will be familiar with INCA and may well have used INCA data in the course of your duties. Indeed, some of you may still be using INCA to this day, as several Police Forces continue to derive benefit from INCA units built in 1998, the technology having outlasted the vehicles it has been installed in several times over! One of the key features of ARTEMIS is the ability to monitor driver behaviour and raise alerts if driver Figure 1 Figure 2 : A 2015 XDR unit pictured alongside the battle-scarred 1998 INCA it recently replaced. Brief Description of IDR Functionality The IDR module in XDR continually monitors data from internal sensors at up to 200Hz and detects when acceleration values exceed pre-configured thresholds. When this occurs the XDR initiates the recording of an IDR file. Page 4 Impact : Spring 2016

2 Features of the IDR functionality include : - A separate IDR file recorded for each incident Each file is 45 seconds duration - 30 seconds prior to impact - 15 seconds post impact Capacitors provide back-up power if vehicle power lost Activation notification to driver via Driver Feedback LED Over-air notification of incident to back office Manual activation via dash button (eg. recording a near miss) IDR files stored on internal Micro SD Capacity monitored and reported over air - no data loss Regular self-tests identify and report any issues over-air XDR IDR files contain the following data elements - Driver ID Vehicle Identification Number (VIN) Device information (serial numbers and software versions) Digital input / output status changes Vehicle CANbus data (varying rates depending upon data type) 10Hz GPS data (date time, position, speed, heading, number of satellites, DOP values) 10Hz OBDII information (vehicle speed, engine RPM, faults) 200Hz 3-Axis accelerometer data - 2 accelerometers Low G and High G - Longitudinal, lateral and vertical accel era tion 200Hz 3-Axis gyroscope data Vehicle roll, pitch and yaw (Note : At Crash Day 2014 the GPS was logged at 1Hz and sensor data at 100Hz) In addition to the above, the XDR offers Journey Data Recording (JDR) functionality. With JDR enabled the XDR stores a second-by-second record of every journey that the vehicle makes. JDR files are written to the internal Micro SD card in the same manner as IDR files and with 32GB storage available there is sufficient capacity to store 2 to 3 years of historical journey information. Having JDR data available in addition to IDR data gives greater context to incident investigation it allows the events leading up to and following an incident to be analysed, not just the 45 second period around the incident itself. Detecting the incident and recording the data is only half the story. APD s Collision Data Analyser (CDA) application reads the captured data, analyses it and presents it in several different visualisations. CDA facilitates complete end-to-end management of journey and incident data, from sensor test and installation record management to full forensic collision data analysis and reconstruction, thereby offering a comprehensive ARTEMIS IDR package. CDA provides the following features and functionality : - Unit sensor and installation tests Calibration Certificate management Data library management Viewing of journey and incident locations on a variety of mapping platforms Graphical analysis of accelerometer, gyro, GPS, OBD/CANbus and i/o data Measurements and calculations of events over distance and time 1-D, 2-D & 3-D reconstruction Comprehensive data import and export capability Integrated video playback Screen capture and recording to assist with report compilation Crash Day 2014 APD played a key role in capturing data for Crash Day 2014, and the engineering team planned the day carefully to ensure that all data was captured. There would be no chances to re-run tests if the kit was accidentally not switched on! Two days before the event, the van was loaded up with as much kit as it could carry equipment, back-up equipment, spare back-up equipment, and just to be on the safe side, back-up spare back-up equipment! We had hot weather gear, cold weather gear, wet weather gear, windy weather gear, there were even mountain bikes and camping gear thrown in. Nothing was left to chance! Prep Day arrived and it was time to install the kit into the vehicles, test it and calibrate it. As there was no access to an external power source in the vehicles we were reliant on trusty old extra long-life alkaline AA battery power. The vehicles would be left overnight so before we departed Bruntingthorpe that evening each vehicle was checked, double checked and treble checked to ensure that the battery packs had been switched off (we had a few spares of course but it would be inconvenient to have to start swapping out battery packs on the day!). Impact : Spring 2016 Page 5

3 Crash Day was then upon us everyone knew their role, everyone knew who they were providing cover for, and we had rehearsed the day many times in advance. At any given time there would be at least one of us up at the start line, where the bullet vehicles began their last journey, to ensure that kit was powered up and functioning. There would be at least 2 people videoing each crash and a further 2 responsible for the GPS surveys (before and after each crash) and also to retrieve the kit from the vehicles not always an easy task after some of the collisions! The retrieved kit was then hurriedly delivered to Crash HQ (the back of the van which had been turned into a mobile command centre with chairs, tables, laptops and even a printer) where one of us was held captive for the entire day downloading and analysing the data from each collision immediately after it happened and ensuring that everything was catalogued and backed up on a regular basis. Aside from an intermittent cellular coverage issue with the GPS survey equipment, everything went pretty much according to plan and the day was a great success. Summary of Crash Day Event organised by ITAI at Bruntingthorpe on 26 th June 2014 Sponsored by APD Opportunity to capture real crash data for validation 7 Hard object collisions (vehicle vs vehicle) 6 Soft object collisions (vehicle vs pedestrian/ cyclist) 2 XDRi units installed in each vehicle - 1 recording IDR data continuously - 1 to test trigger algorithms and IDR file creation High speed video (400 fps) of each crash GPS survey data of each crash Analysis of Crash 13 Zafira vs Ambulance Preparation Two XDRi units were installed in each vehicle, one continually logging data at 100Hz and one set to trigger as a result of the collision and recording a 45 second IDR file. For each hard collision there were four data files recorded. In addition to this, high speed video from at least 2 angles was recorded and accurate GPS positions using survey grade equipment were logged. For the target vehicle the pre and post collision positions were logged. The purpose of the GPS survey data and high speed video was to provide an external reference enabling comparisons with the IDR data to be made in order to corroborate the accuracy of the calculations made by CDA from the IDR files. Under normal circumstances the XDRi would have had a connection to the vehicle OBDII and/or CANbus in order to receive additional data such as speed and engine RPM. This was not possible on Crash Day because the vehicles were drained of fuel and did not have their engines running. Analysis Firstly we look to calibrate the sensors and determine the accuracy of the GPS receiver. CDA provides visual indication of the data and allows it to be calibrated to known reference points. For the bullet vehicle the bottom graph shows that the GPS accuracy is good with a confidence factor over 0.95 indicating an accuracy of the speed reading of around 0.1 mph. This is what we would expect given the clear conditions on the day. The top graph shows the accelerometer calibrated against the GPS speed trace and shows an almost perfect match up to the point of the collision. This is what we would expect given the flat nature of the test track. After the collision we can see that the GPS overruns due to the sudden change in speed, whereas the sensor tracks the real speed. Note : Only one of the 13 collisions staged at Crash Day 2014 is considered in detail in this article. Material relating to the other collisions will shortly be available in the Members section of the Institute website. What follows is an in-depth walk-through of the process of investigation. Figure 3 : Crash diary Page 6 Impact : Spring 2016

4 Figure 4 Using this calibrated data the Event Analysis shows : - Impact speed of 20.1 m/s (45.0mph +-0.1) from the combined GPS and sensor A peak deceleration of -224 m/s² (-22.8g) Speed change during the impact phase of 16.4 m/s (-36.7 mph) Residual speed as the vehicle finally came to rest over the following second Figure 5 The gyroscope pitch sensor registered a dip in the vehicle nose peaking at 140 /s which accounts partly for why not all of the deceleration was registered by the accelerometer longitudinal sensor. The Gyroscope is an essential tool in an IDR in understanding the correct motion of the vehicle. Impact : Spring 2016 Page 7

5 Figure 6 The graph shows the bounce in the vehicle s suspension. Figure 7 Putting the data in context we can observe the incident on the map and generate from CDA a 3D reconstruction at the point of impact which shows the nose of the vehicle pushed down to the surface of the road. Page 8 Impact : Spring 2016

6 Figure 8 The 3D reconstruction in CDA quickly provides the investigator with insight into the nature of the collision and of the 30 seconds of driving leading to up to it. If this replay is performed at the collision scene it can help the investigating officer to capture further evidence before it is lost as the road is cleared. In the absence of a Zafira template, and alternative saloon vehicle template was used for 3D modelling purposes. Figure 9 Turning our attention to the hapless ambulance, inspection of the accelerometer and gyro data shows an initial sharp roll to the left followed by a larger roll to the right. Impact : Spring 2016 Page 9

7 Figure 10 Note the precision with which the modelled version of the target vehicle is shown, when compared with the video film taken at the time of the tests. In a sequence of screenshots and photographs set out above at Figure 10, 3D modelling based on the captured accelerometer and gyro data shows the target vehicle (the ambulance) in an identical configuration to that recorded on the video film taken at the time of the tests. In short, the video film confirms and validates the analysis. The ARTEMIS Incident Data Recorder provides a comprehensive set of tools for the collision investigator. It provides the complex numerical analysis required to perform calculations in 1D and 2D, and provides visualisation tools for understanding the vehicle motion in 3D. The ITAI 2014 event provided a means to validate the system. Contact : Martin Worrell address martin.worrell@apdcomms.com Technical Director APD Communications Telephone Page 10 Impact : Spring 2016