Reconstruction of a Crash Using Exemplar EDR Records

Introduction

In this case study, we present an example of incorporating publicly available, exemplar data from Event Data Recorders (EDRs) into an accident reconstruction. In current accident reconstruction practice, it is commonplace to have objective, measured data from at least one of the involved vehicles. This data, which is recorded by on-board vehicle systems, is often obtained with the Crash Data Retrieval (CDR) system. One might imagine that the increasing amount of data accessible from vehicles via this system would result in clarity about what happened in the crash and a diminished role for accident reconstructionists. However, there are many instances when an accident reconstructionists is needed to interpret the data. Oftentimes, the correct interpretation of the data from the vehicle is not apparent on the surface and someone needs to analyze its meaning within the context of the crash.

Early CDR reports were short, often with a page or two of data, and were relatively easy to digest. Now, it is not uncommon for CDR reports to be over 60 pages long. In many cases, this data is not laid out in a way that enables a lay person to read the report and understand what happened in the crash. Without a technical background in accident reconstruction, it can be difficult to know which recorded data is related to the subject crash, which data useful, what limitations exist in the data, what specifically was happening to the vehicle when the data was collected, and where and when during the sequence specific data was collected. When these questions arise, it can sometimes be helpful to examine not only the CDR report from the vehicle involved in the crash, but also CDR reports from similar make and model vehicles involved in other crashes.

Examination of exemplar CDR reports can sometimes be carried out through the National Highway Traffic Safety Administration’s (NHTSA’s) Crash Investigation Sampling System (CISS).[1] According to NHTSA’s website,[2] “CISS collects detailed crash data to help scientists and engineers analyze motor vehicle crashes and injuries. CISS collects data on a representative sample of minor, serious, and fatal crashes involving at least one passenger vehicle – cars, light trucks, sport utility vehicles, and vans – towed from the scene. After a crash has been sampled, trained Crash Technicians obtain data from crash sites by documenting scene evidence such as skid marks, fluid spills, and struck objects. They locate the vehicles involved, document the crash damage, and identify interior components that were contacted by the occupants. On-site inspections are followed-up with confidential interviews of the crash victims and a review of medical records for injuries sustained in the crash. CISS uses emerging technologies and methods to acquire quality data.” Often the data collected by CISS investigators includes event data from the involved vehicles’ airbag control modules (ACMs).

Background on the Subject Crash

The subject collision involved a 1997 Ford F250 (Vehicle 1) striking the rear of a 2018 Chevrolet Equinox (Vehicle 2) and pushing it into the rear of a 2003 Hyundai Accent (Vehicle 3). According to the Traffic Collision Report (TCR), Driver 3 was slowing or stopped with the intention of making a left turn. Driver 2 was slowing behind Vehicle 3. Driver 1 failed to respond sufficiently to the slowing or stopped vehicles ahead and collided with the rear of Vehicle 2. This crash occurred under daylight conditions and on a dry road. Vehicle 1 deposited 65 feet of skid marks prior to collision.

Driver 2 testified that she was behind another vehicle which intended to make a left-hand turn. She said the vehicle ahead of her did not come to an abrupt stop, but a slow stop which caused her to “lightly” apply her brakes. She came to a complete stop about half a vehicle-length behind the vehicle ahead of her. She first became aware of the vehicle behind her when she heard screeching and looked in her rearview mirror. She did not know how far back the vehicle was at that moment, but she said she panicked and applied her brakes firmly prior to the collision. Her car was not totaled and was repaired, and she was still in possession of the vehicle. Another accident reconstructionist imaged data from the ACM on this vehicle, and that data was provided to us.

The photograph in Figure 1 shows the damage to the rear of the 2018 Chevrolet Equinox from the first collision. The photograph in Figure 2 shows the damage to the front of the Chevrolet from the second collision. The damage to the front was limited to the front driver side quarter panel and red paint transfer on the bumper. This all-wheel drive vehicle had the Premier trim package and was equipped with a 2.0-liter Turbo DOHC 4-cylinder gasoline engine, a 9-speed automatic transmission, and antilock brakes (ABS). The Chevrolet weighed approximately 3,800 pounds at the time of the collision, including the driver. This vehicle was equipped with an EDR in the airbag control module (ACM, 01049_SDM30-autoliv_r015), and I was provided with the download of data from this module. The airbags in the Chevrolet did not deploy and the CDR report stated that no collision events were recovered.

Figure 1 – Rear of the Chevrolet After the Collision

Figure 2 – Front of Chevrolet after the Collision

Figure 3 shows the damage to the front of the 1997 Ford F250. The front bumper exhibited mild deformation, as did the leading edge of the hood. This rear-wheel-drive vehicle was equipped with the LD SuperCab trim, a 5.8-liter V8 gasoline engine, and a 4-speed automatic transmission. This vehicle was not equipped with an event data recorder (EDR). The Ford weighed approximately 4,900 pounds at the time of the crash, including the driver. Figure 3 shows the damage to the front of the 1997 Ford F250. The front bumper exhibited mild deformation, as did the leading edge of the hood. This rear-wheel-drive vehicle was equipped with the LD SuperCab trim, a 5.8-liter V8 gasoline engine, and a 4-speed automatic transmission. This vehicle was not equipped with an event data recorder (EDR). The Ford weighed approximately 4,900 pounds at the time of the crash, including the driver.

Figure 3 – Front-End Damage to the Ford

Collision Speed and Severity Analysis

The subject collision involved the Ford F250 colliding with the rear of the Chevrolet Equinox. This initial collision then pushed the Chevrolet into the rear of the Hyundai. To determine the speed of the Ford F-250 when it collided with the rear of the 2018 Chevrolet Equinox, we first assessed the reason that the ACM on the Chevrolet did not contain an event either from the initial collision with the Ford or from the subsequent collision with the Hyundai. There were several possibilities:

1)      The EDR on the Chevrolet was not capable of recording rear impacts.

2)      An event was initially recorded but was later erased by the system.

3)      Neither collision was severe enough to trigger an event being recorded.

Code of Federal Regulations (CFR), Title 49, Volume 6, Part 563 “specifies uniform, national requirements for vehicles equipped with event data recorders (EDRs) concerning the collection, storage, and retrievability of onboard motor vehicle crash event data.” This regulation became effective for vehicles manufactured on or after September 1, 2012, so it was applicable to the Chevrolet Equinox involved in the subject collision. This regulation requires vehicles to record events when the change in velocity (Delta-V) experienced by the vehicle exceeds 8 km/h (5 mph) within a 150-millisecond interval. Thus, the Chevrolet should have recorded an event if the vehicle experienced a forward or rearward change in velocity greater than 5 mph.

For older General Motors ACMs (Generation 1 through 3), prior to the implementation of CFR Part 563, non-deployment events were erased by the system after 250 ignition cycles (about 1 month of normal driving). Since 2012, this is no longer the case. Thus, the likely reason that an event was not recorded on the Chevrolet is that the Delta-Vs it experienced from interacting with the Ford and the Hyundai were both less than 5 mph.

To further evaluate the reason the Equinox did not report an event from the subject collision, we obtained similar crash cases from the NHTSA’s CISS database. The opening query screen for this database is depicted in Figure 4. We queried data for Chevrolet Equinoxes of Model Years 2016 through 2020 that were involved in rear impacts. We also requested cases that included EDR records. This query returned 11 cases (Figure 5), and we chose 5 of these collisions as suitable exemplar cases. These cases are discussed next.

Figure 4 – Opening Query Screen for CISS Database

Figure 5 – CISS Database Query Results

CISS Case #1-19-2020-149-02

This first case involved the front of a 1999 Nissan Frontier pickup (Figure 6) colliding with the rear of a 2016 Chevrolet Equinox (Figure 7). The Chevrolet experienced a subsequent impact to the front right (Figure 8). The Bosch Crash Data Retrieval (CDR) system was used to download data from the ACM on the Chevrolet, and the ACM (01042_SDM10P-autoliv_r021) reported a non-deployment event. The ignition cycles at the time of the event were 10,350, and at the time of the download of the ACM they were 10,354. Thus, the non-deployment event was likely from the collision under investigation by the CISS investigators. The CDR report indicated that the Chevrolet Equinox in this collision experienced a positive (rear impact) change in velocity (Delta-V) of approximately 12 mph in 96 milliseconds from the rear collision (Figure 9). The vehicle experienced a peak longitudinal acceleration during the collision of 13g.

Figure 6 – Nissan Frontier from CISS Case #1-19-2020-149-02

Figure 7 – Rear of Chevrolet Equinox from CISS Case #1-19-2020-149-02

Figure 8 – Front of Chevrolet Equinox from CISS Case #1-19-2020-149-02

Figure 9 – Longitudinal Delta-V Data from the CDR Report for CISS Case #1-19-2020-149-02

This collision event recorded by the ACM is, first, an indication that the ACMs on Chevrolet Equinoxes are capable of recording rear impacts. Thus, that cannot be the reason that no event was recorded in the collisions involving the subject Equinox. Second, this exemplar case is an indication that a Delta-V of 12 mph is greater than the threshold for an event to be recorded. This is consistent with the requirements of CFR Part 563.

In addition, according to the Data Limitations section in the CDR report for this CISS case: “There are two types of recorded crash events for Front, Side, and Rear (FSR) Events. The first is the Non-Deployment Event. A Non-Deployment Event records data but does not deploy the air bag(s). The minimum SDM Recorded Vehicle Velocity Change, that is needed to record a Non-Deployment Event, is five MPH [8 km/h]. A Non-Deployment Event contains Pre-Crash and Crash data. The oldest Non-Deployment event can be overwritten by a Deployment Event, if all three records are full and the Non- Deployment Event is not locked. A Non-Deployment Event can be overwritten by a more recent Non-Deployment Event if all three records are full, and the Non-Deployment is older than approximately 250 ignition cycles…The second type of SDM recorded crash event for FSR Events is the Deployment Event. It also contains Pre-Crash and Crash data. Deployment Events cannot be overwritten or cleared by the SDM…The SDM can store up to three Events.” These statements from the data limitations confirm that rear events are recorded, and that an event will be recorded if it exceeds a change in velocity of 5 mph. A non-deployment event can be overwritten, but only if all of the three available slots for events are full. Since there were no other events reported by the ACM on the subject Equinox, we can conclude that there was not an event initially recorded but later overwritten, and that the vehicle did not experience a change in velocity exceeding 5 mph in the subject collision.

CISS Case #1-24-2019-133-03

This next case involved the front of a 2008 Honda Civic (Figure 10) colliding with the rear of a 2019 Chevrolet Equinox (Figure 11). The Chevrolet was then pushed forward into the rear of a 2013 Honda Civic. The front-end damage to the Chevrolet from the second collision is depicted in Figure 12. The Bosch Crash Data Retrieval (CDR) system was used to download data from the ACM on the Chevrolet, and the ACM (01049_SDM30-autoliv_r015) reported a non-deployment event. This event occurred at 1,303 ignition cycles, and the download of the ACM occurred at 1,310 ignition cycles. Thus, the downloaded event was from the collision under investigation. The CDR report indicated that the Chevrolet Equinox in this collision experienced a positive change in velocity (Delta-V) of approximately 14 mph in 146 milliseconds from the rear collision (Figure 13). The vehicle experienced a peak longitudinal acceleration during the rear collision of 13.0g. This case and the data limitations listed in this CDR report confirm the conclusions from the prior CISS case.

Figure 10 – Honda Civic from CISS Case #1-24-2019-133-03

Figure 11 – Rear of Chevrolet Equinox from CISS Case #1-24-2019-133-03

Figure 12 – Front of Chevrolet Equinox from CISS Case #1-24-2019-133-03

Figure 13 – Longitudinal Delta-V Data from the CDR Report for CISS Case #1-24-2019-133-03

CISS Case #1-25-2020-119-03

This case involved the front of a 2010 Toyota Scion tC colliding with the rear of a 2017 Chevrolet Equinox (Figure 14). The Chevrolet was then pushed forward into the rear of a 2015 Chevrolet Traverse, and then the Traverse was pushed forward into the back of a 2007 BMW 3-series. The front-end damage to the Chevrolet Equinox from the second collision is depicted in Figure 15. The Bosch Crash Data Retrieval (CDR) system was used to download data from the ACM on the Chevrolet Equinox, and the ACM (01042_SDM10P-autoliv_r021) reported two non-deployment events. The download of these events occurred at 5,289 ignition cycles and both non-deployment events occurred at 5,275 ignition cycles. These events were likely from the collision under investigation. The first non-deployment event was from the rear collision and the second was from the subsequent front collision. The CDR report indicated that the Equinox first experienced a positive Delta-V of approximately 9 mph in 158 milliseconds from the rear collision (Figure 16) and then a negative Delta-V of approximately 7 mph in 134 milliseconds (Figure 17). During the first collision, the Equinox experienced a maximum longitudinal acceleration of 9.8g, and during the second collision, the Equinox experienced a maximum longitudinal acceleration of -5g. This case and the data limitations listed in this CDR report confirm the conclusions from the prior CISS case.

Figure 14 – Rear of Chevrolet Equinox from CISS Case #1-25-2020-119-03

Figure 15 – Front of Chevrolet Equinox from CISS Case #1-25-2020-119-03

Figure 16 – Rear Collision Delta-V Data from the CDR Report for CISS Case #1-25-2020-119-03

Figure 17 – Front Collision Delta-V Data from the CDR Report for CISS Case #1-25-2020-119-03

CISS Case #1-29-2019-112-03

This case involved the front of a 2009 Volkswagen Jetta (Figure 18) colliding with the rear of a 2019 Chevrolet Equinox (Figure 19). The Bosch Crash Data Retrieval (CDR) system was used to download data from the ACM (01049_SDM30-autoliv_r015) on the Chevrolet, and the ACM reported two non-deployment events. The download of the ACM occurred at 1,359 ignition cycles, the first event occurred at 470 ignition cycles, and the second event occurred at 1,341 ignition cycles. Thus, the first non-deployment event was from a prior collision, with a reported Delta-V of positive 5 mph in 108 milliseconds (Figure 20). This confirms that a 5 mph Delta-V will be recorded by the ACM on a Chevrolet Equinox. The second non-deployment event was from the collision being documented by CISS investigators, with a reported Delta-V of positive 6 mph in 132 milliseconds. The prior collision produced a maximum longitudinal acceleration of 5g, and the subject collision produced a maximum longitudinal acceleration of 4.2g. This case and the data limitations listed in this CDR report confirm the conclusions from the prior CISS case.

Figure 18 – Volkswagen Jetta from CISS Case #1-29-2019-112-03

Figure 19 – Chevrolet Equinox from CISS Case #1-29-2019-112-03

Figure 20 – Prior Rear Collision Event for the Chevrolet from the CDR Report for CISS Case #1-29-2019-112-03

Figure 21 – Rear Collision Delta-V Data for the Chevrolet from the CDR Report for CISS Case #1-29-2019-112-03

CISS Case #1-30-2020-143-02

This next case involved the front of a 2019 Dodge Ram Promaster colliding with the rear of a 2020 Chevrolet Equinox (Figure 22). The Chevrolet was then pushed forward into the rear of a 2012 Jeep. The front-end damage to the Chevrolet from the second collision is depicted in Figure 23. The Bosch Crash Data Retrieval (CDR) system was used to download data from the ACM on the Chevrolet, and the ACM (01049_SDM30-autoliv_r015) reported two non-deployment events. These events occurred at 1,925 ignition cycles, and the download of the ACM occurred at 1,943 ignition cycles. The downloaded events were likely from the collision under investigation. The CDR report indicated that the Chevrolet Equinox in this collision first experienced a positive Delta-V of approximately 13 mph in 126 milliseconds from the rear collision (Figure 24). The vehicle experienced a peak longitudinal acceleration during this rear collision of 11.0 g. The Equinox next experienced a frontal impact with a change in velocity of approximately 7 mph occurring at 124 milliseconds. This collision produced a maximum longitudinal acceleration of -8.2 g. This case and the data limitations listed in this CDR report confirm the conclusions from the prior CISS case.

Figure 22 – Rear of Chevrolet Equinox from CISS Case #1-30-2020-143-02

Figure 23 – Front of Chevrolet Equinox from CISS Case #1-30-2020-143-02

Figure 24 – Rear Collision Event for the Chevrolet from the CDR Report for CISS Case #1-30-2020-143-02

Figure 25 – Front Collision Event for the Chevrolet from the CDR Report for CIS Case #1-30-2020-143-02

Momentum and Pre-Impact Braking Analysis

Based on the requirements of Part 563, on the CISS exemplar case reports, and using principles of physics, we concluded that when the Ford F-250 collided with the Chevrolet Equinox, the Equinox experienced a positive change in velocity of 5 mph or less. During this collision, the Equinox experienced a peak longitudinal acceleration of 5 g or less. Using the principle of conservation of momentum, and accounting for tire forces acting during the collision, we concluded that the Ford F-250 was traveling 17 mph or less when it contacted the rear of the stopped Chevrolet Equinox. When it was pushed forward, the Chevrolet Equinox then contacted the Hyundai at a speed of 5 mph or less. This produced a rearward change in velocity for the Chevrolet Equinox of 2 mph or less. Investigating officers documented 65 feet of pre-collision skidding from the Ford F-250. Assuming a deceleration during skidding of 0.7 to 0.8 g, and accounting for the time required to build-up this deceleration, we calculated the Ford was traveling 45 mph or less at the time the brake application began.

An important point here is in relationship to what we did NOT do with the exemplar CISS cases. We did not do any visual comparison between the damage to the Chevrolet Equinoxes in the CISS cases and the subject case and attempt to use that as a means to estimate the severity of the subject collision. To understand why, consider the following statement from an article titled “Crash Pulse and Delta V Comparison in a Series of Crash Tests with Similar Damage (BEV, EES),” written by Ronald Woolley and Alan Asay and published by the Society of Automotive Engineers in 2008 (Paper Number 2008-01-0168): “Observed damage to a vehicle involved in an accident tells an incomplete story…Knowledge of the collision partner is essential in order to complete an accident reconstruction…In general, it has been shown that in a two vehicle crash the damage to each vehicle is related to the mass ratio and stiffness ratio of the collision partners as well as to the closing speed.” In other words, the damage that resulted to the Chevrolet Equinoxes in the CISS exemplar cases was the result of not only the closing speed between the two vehicles, but also the mass and stiffness ratios of the involved vehicles. Unless the mass and stiffness ratios in the exemplar cases were the same as the mass and stiffness ratios in the subject collision, no direct comparison of the damage can be made.

Summary of Conclusions

1.      The Chevrolet Equinox was reportedly stopped when it was impacted by the Ford F-250.

2.      When he began applying his brakes, the driver of the Ford was traveling 45 mph or less.

3.      Prior to this collision, the Ford F-250 deposited 65 feet of skid marks as a result of the driver’s attempt to avoid the collision by braking.

4.      When it collided with the rear of the Chevrolet Equinox, the Ford F-250 was traveling 17 mph or less.

5.      The Chevrolet Equinox experienced a positive (rear impact) change in velocity of less than 5 mph as a result of the collision from the Ford F-250.

6.      The Chevrolet Equinox experienced a negative (front impact) change in velocity of less than 2 mph as a result of the collision with the Hyundai.

[1] https://crashviewer.nhtsa.dot.gov/CISS/SearchFilter#

[2] https://www.nhtsa.gov/crash-data-systems/crash-investigation-sampling-system