The Wild West days of the automotive event data recorder (EDR) are drawing to a close. As of Sept. 1, 2012, all light vehicles manufactured in the model year 2013 and beyond must adhere to new data recording and retention standards established by the National Highway Transportation Safety Administration (NHTSA) under 49 C.F.R. 563.1-563.12.
Simply put, an EDR is a microprocessor tasked with recording data. In the case of an airbag EDR, commonly referred to as a black box, the mission is to collect crash data. The data recorder embedded in airbag restraint systems is the only vehicle EDR affected by the NHTSA rule.
When EDRs stepped into the limelight in 1994, fewer than 10 percent of vehicles had them on board. That figure had changed significantly by 2005, when NHTSA published statistics showing 64 percent of new models with EDRs. Today that figure exceeds 80 percent.
Once in effect, the new rule will require automakers to disclose all onboard safety-related data recording functions to consumers. It also sets data collection and reporting standards if manufacturers choose to install EDRs. Even without a forced mandate to embed EDRs in their airbags, automakers were encouraged by regulators to adopt the technology. The real-time crash data that manufacturers could tap into would help them reduce costs, design safer vehicles, and, by extension, sell more cars.
Consumer prospects are more mixed. While they will be notified of the presence of data recorders in their airbag safety systems, car buyers will not be able to remove them without nullifying warranties. As for those who prefer EDR-free vehicles, they will be selecting from a rapidly shrinking list of new cars because the latest models are essentially rolling microprocessor networks.
The lack of drama surrounding the activation of NHTSA’s rule speaks to the generalized acceptance of a technology that initially provoked anger from a fearful public who saw its use in the passenger fleet as an assault on privacy rights. Ten-plus years of debate and high-tech advances have recast the automotive EDR as a benign form of data gathering, especially in light of more intrusive surveillance like GPS and driver-monitoring products.
EDR Origins
In the late 1980s, automakers recognized that they could record the information being monitored by their airbag control modules and use it as real-life crash data to improve their vehicle safety systems. These early EDRs fulfilled a troubleshooting role much the way aviation black boxes that record flight parameters and cockpit activity do as a means of dissecting air disasters. The differences between the automotive and aviation models were significant at that time and remain so today. Automotive EDRs record very limited data sets and are more vulnerable to damage than aviation black boxes. In addition, an airbag EDR is not a passenger compartment spy; it does not have the capability to monitor or record conversations and behavior or track vehicle location.
A Look Under the Hood
The EDR relies on sensors distributed throughout the vehicle to monitor systems in anticipation of an impact. The system wakes up when it suspects an imminent accident—usually when the vehicle experiences some kind of minor impact, perhaps as insignificant as taking a speed bump too fast or hitting a pothole. Typically, deployment data gets locked in permanently, while near-deployment data is stored until supplanted by another event. A quick comparison of early EDRs versus today’s models shows a significant increase in the scope and amount of data recorded by the newer generation (see Chart 1 below).
Chart 1: The Evolution of EDR Technology
-
Speed
-
RPM
-
Percentage throttle
-
Brake status
-
Front/rear Delta V
-
Seatbelt status
-
Ignition cycle count
-
1 non-deployment
-
1 deployment
-
5 seconds pre-crash
-
Rate: once per second
Current Generation EDR
EDR Vulnerability—Crash Data Retrieval
Tapping into and interpreting an EDR’s crash data has been among the most contentious aspects of its use. For years, automakers considered the module to be proprietary, so they made scant effort to standardize crash data retrieval (CDR), data gathering, and reporting. Once the technology proved useful to a broader audience, the auto industry took notice and started easing access to the module’s data stream. Initially, only Ford, General Motors, and Chrysler licensed a commercial interface. All other data downloads had to be initiated through the specific automaker—often a Herculean task. Even today with the Big Three models, there is little commonality in the data collected, sample rate, recording period, communications protocols, or module connectors. This inconvenient reality applies as well to vehicles within a single fleet. Each model, even different model years of the same vehicle, can have unique EDR versions. On the upside, the differences will diminish as the technology matures under NHTSA’s rules.
Despite the slow progress of commercial CDR services, there is significant change afoot. Toyota’s recent public relations disaster of unintended acceleration allegations has spurred compliance with the NHTSA rule. Both Toyota and Honda have recently joined Ford, GM, and Chrysler and made their data downloadable with the same interface. Time will tell how the other automakers will adhere to the new data retrieval standards—new CDR vendors could appear, and manufacturers might elect to develop their own versions.
Limits of Usability
It is easy to assume that, once functionality and data collection/retrieval methods improve across all vehicle makes and models, EDRs will be able to determine accident causation in all cases and assign fault. That view is, however, unrealistic. For now, the data being recorded is insufficient for detailed accident reconstruction. An EDR delivers information about vehicle system operations at a specific point in time but does not reveal anything about its location on the road, degree of driver distraction, steering input, or damage. Even if an EDR could see all, there are many crash scenarios beyond its capabilities, like tire failures, sideswipes, and phantom vehicles.
In addition, there are plenty of examples where the EDR data can be misleading. Imagine a scenario where the download reveals that the driver’s seatbelt was buckled, but the driver’s severe injuries are inconsistent with seatbelt use. In this case, the crash data could not reveal that the driver elected to sit on his buckled seatbelt rather than use it properly.
The bottom line for EDR use is that the downloaded data is usually helpful, but it is best used as an adjunct to a thorough engineering analysis and not as stand-alone testimony. The analysis and interpretation of the data are the realm of automotive engineers who are aware of the technology’s limitations and are trained in a full range of analytical methods.
There are also caveats for crash data retrieval that are crucial for minimizing chain-of-custody questions and spoliation claims. From the top, tampering with airbag sensors or attempting to remove an EDR from a vehicle can hinder airbag operation and void warranties. Plus, airbags are explosive devices that can be dangerous when handled improperly. As for the module itself, jolting or mishandling the module when powering up for a download can trigger data overwrites or loss. For these reasons and more, especially in severe crashes, all interested parties should be present to observe downloads. The ideal strategy post-accident is to get the data promptly before someone attempts to start the vehicle.
The Way Forward
There is some irony in the intense scrutiny afforded one small piece of onboard technology. Nowadays, an airbag EDR is only one example of a vehicular data-gathering module—the newest-model cars can have 300 or more microprocessors. While most will never be relevant to accident investigations, some might assume that role eventually. Until then, an airbag EDR remains the most useful module for accident investigations; thus, the standardization by NHTSA is necessary. However, those who use the technology need to be aware that greater changes are afoot. The future issues of automotive technology lie more in the implications of broadband connectivity and the possibility of driverless vehicles rather than in the regulated mission of one notorious microprocessor.
Peter R. Thom, a CLM fellow since 2008, is president of Peter R. Thom and Associates, Inc., a national firm of consulting automotive engineers. Gregory J. Quan is the firm’s managing engineer. They can be reached at (800) 874-1664, www.prtassoc.com.
The Framework for EDR Use in Legal Actions
The primary legal issues have revolved around ownership of the EDR data and the extent to which Fourth Amendment privacy protections prohibit seizure of that data absent consent.
To date, 13 states have enacted laws regarding access to EDR data. All require that the vehicle owner consent to a data download. Several of these statutes also preserve original-owner rights to consent when the insurer takes salvage ownership. Even if a state lacks EDR-specific statutes, computer trespass laws may apply to prohibit access without owner consent.
Regarding courtroom use, EDR evidence has cleared both the Frye and Daubert thresholds of admissibility. In criminal cases, EDR data has generally been used to prosecute drivers who drove recklessly before fatal collisions. In civil actions, it has primarily been used to prove or rebut vehicle-defect claims.
Many of the criminal cases have addressed the question of whether or not EDR data can be seized without a search warrant. At issue is whether or not the defendant has a reasonable expectation of privacy in the data so as to invoke the protections of the Fourth Amendment. To date there is no consensus among the courts. Clearly the law in this area will evolve as more vehicles are equipped with EDRs and as the modules are modified to capture a broader scope of information.
EDRs in Action
Wondering how EDRs can affect your auto claims? Here are two examples of real claims that relied on crash data imaged by Crash Data Group and its Bosch crash data retrieval tool.
What the Insured Claimed
The vehicle in question was purchased four days prior to a claim and was damaged before a pre-inspection was completed. The insured claimed the car was damaged while parked. No one else was in possession of the keys. The insured drove the vehicle to the police station to file a report.
Result
The insured signed a release to image data, allowing the data to be retrieved. The vehicle was examined at the salvage yard. The image of the EDR data was retrieved and a data analysis report was completed. It determined that the vehicle was not damaged while parked. Rather, it was hit from behind while slowing. The insured was asked to provide an EUO; however, they never responded.
What the Insured Claimed
The insured claimed that he parked his car overnight in a large retailer’s parking lot. When he returned the next day, he discovered that the car was damaged while parked. No one else was in possession of the keys to the car. The insured had the car towed to the repair shop and the repairs were completed before it was inspected. Therefore, there were no damaged parts to inspect.
Result
The insured signed a release to image data, allowing data to be retrieved. The repaired vehicle was examined and the image of the EDR data was retrieved. The data analysis report determined that the vehicle was not damaged while parked. The insured was offered the opportunity to recant. The claim was denied and no payment was made.