Summary
accepting testimony of an expert witness that "there are no organizations within the automotive community which question the reliability of EDR devices"
Summary of this case from State v. ByardOpinion
No. 05-03-00372-I.
Decided November 21, 2005.
Bruce Holmes, Assistant Prosecutor for plaintiff.
Joseph P. Depa, Jr., for defendant.
The State filed a motion for a Rule 104 ( N.J.R.E. 104) hearing regarding the admissibility of automobile "black box" evidence. It is asserted by the State, in this death-by-auto charge, that the vehicle driven by defendant at the time of the collision was equipped with what is commonly known as a "black box," or a computerized data recorder. This "black box" is similar to the data recorder in the aeronautic industry that is often sought after an airplane crash to determine various items of data, to wit; altitude, latitude, speed, and other mechanical operation data within the aircraft. The State seeks to introduce into evidence the data recorded, together with expert witness testimony to prove an element of the crime charged. To prove death by auto under N.J.S.A. 2C:11-5, the State must establish the following elements:
1) Defendant was driving an automobile, (a "means of conveyance propelled otherwise than by muscular power"
2) The motor vehicle caused the death of the victim, and
3) Defendant operated his vehicle recklessly.
It is undisputed that the vehicle in question was an automobile and was operated by defendant, and that on September 14, 2004, the vehicle, a 2003 Hummer H2, ran over a Honda Civic from the rear, while the victim was stopped for a red light on an off-ramp of Route 278. The victim was pronounced dead at the scene. To prove the third element of the crime, recklessness, the State seeks to introduce into evidence the data retrieved from the "black box."
The prosecution, pursuant to a warrant issued on November 17, 2004, seized the "black box" in defendant's vehicle and had the recorded data retrieved by a New Jersey State Trooper, utilizing download equipment. The data downloaded purported to show that defendant's vehicle was operating at seventy-six miles per hour within five seconds prior to the accident and at sixty-seven miles per hour at the time of impact. Additionally, the data purported to show that the brakes were not applied until zero-to-one second before impact, and that there were no mechanical defects in the vehicle. It is the State's position that this scientific evidence is sufficiently reliable and trustworthy to allow it to be presented, through expert testimony, to a jury. This is a case of first impression in New Jersey regarding the use of this scientific data against a defendant in a criminal case.
The Rule 104 hearing was held on October 21, 2005, at which time the State presented one witness, W.R. Haight, as an expert in accident reconstruction and data recording devices in motor vehicles. For the reasons stated below, the State's motion to permit the use of the "black box" data and expert testimony at time of trial is granted because the State has clearly established, by a high level of proof, the general acceptance of its technology in the automotive field, since it has passed from the experimental to the demonstrable stage.
Haight is the director of the Collision Safety Institute, a crash research, training and consulting center operating primarily in the United States, Canada, and Australia.
Background
Crash-sensing devices are commonly used throughout the automobile industry. They have multiple functions, including: (1) determining if a severe enough impact has occurred to warrant deployment of the air bag; (2) monitoring the air bag's components; and (3) permanently recording information pertaining to the operation of the vehicle. The crash-sensing device on General Motors (GM) vehicles is known as the Sensing and Diagnostic Module (SDM); these devices have been installed on GM vehicles since 1994. The SDM contains software that analyzes the deceleration of a vehicle to determine whether a deployment event (usually a crash) has occurred. When the SDM senses a deployment event, or a deceleration that is not severe enough to deploy an air bag (a near-deployment event), it records that information to the microprocessor's electrically erasable programmable read-only memory (EEPROM). Data that the SDM records includes:
Restraint Control Module (RCM) is the airbag module used in Ford vehicles since 1998.
(1) Vehicle speed five seconds before the deployment or near-deployment event;
(2) Engine speed five seconds before the deployment or near-deployment event;
(3) Brake status five seconds before the deployment or near-deployment event;
(4) Throttle position five seconds before the deployment or near-deployment event;
(5) Engine revolutions (RPM) at the time of the deployment or near deployment event;
(6) Status of the driver's seat belt switch (on or off) at the time of deployment or near-deployment event;
(7) Status of the brake switch (on or off) during the five seconds before the deployment or near-deployment event;
(8) Time from maximum deceleration to vehicle impact;
(9) Time from vehicle impact to air bag deployment;
(10) Maximum deceleration for a non-deployment event;
(11) Diagnostic trouble codes present at deployment;
(12) The time the deployment or near-deployment event occurred.
[ Bachman v. GMC, 332 Ill.App.3d 760 , 267 Ill.Dec. 125, 776 N.E.2d 262 (2002)]
On GM vehicles, the Vetronix Crash Retrieval System (CDR) consists of hardware and software that downloads precrash and crash data from the SDM to a laptop computer. Since mid-2000, the Windows-based CDR software presents this data in easy-to-read graphs and tables. The CDR is known generically as an Event Data Recording device (EDR), or "black box." Rule 702 Analysis
New Jersey Rule of Evidence 702 governs the admission of expert testimony. The Rule provides:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.
[N.J.R.E. 702]
Three requirements must be met before expert testimony can be admitted:
"(1) the intended testimony must concern a subject matter that is beyond the ken of the average juror;
(2) the subject of the testimony must be at a state of the art such that an expert's testimony could be sufficiently reliable; and
(3) the witness must have sufficient expertise to explain the intended testimony."
[ State v. Harvey, 151 N.J. 117 , 169, 699 A.2d 596 (1997); (quoting State v. Kelly, 97 N.J. 178 , 208, 478 A.2d 364 (1984)); N.J.R.E. 702, 1991 Supreme Court Committee Comment]
It is not disputed that the subject matter is well beyond the ken of the average juror. For instance, someone would need to understand both the precision and accuracy of an EDR in order to be knowledgeable about the reliability of the data taken from it. In order to understand precision, someone would need to know the basic operation and data collection capabilities of the device. In order to understand accuracy, one would need to know, among other things, about the sources of potential error in EDR devices, their physical memory limitations, and how the data collected from crash tests differs from data collected from "real world" crashes. One can assume few individuals possess such knowledge.
For example, the General Motors EDR samples longitudinal acceleration every 312 microseconds and integrates the average of four such samples.
The qualifications of the State's expert are not in dispute. He has given expert witness depositions, affidavits and/or trial testimony in Arizona, California, Connecticut, Georgia, Michigan, Missouri, New York, Pennsylvania, Texas, Washington, Australia and Singapore. He is also one of two experts in the country certified to teach classes on data recorders placed into newer vehicles. Haight appears to have sufficient expertise to explain the intended testimony.
The remaining factor is whether EDR technology is at a state of the art such that an expert's testimony could be sufficiently reliable to present to the jury. N.J.R.E. 703 provides:
The facts or data in the particular case upon which an expert bases an opinion or inference may be those perceived by or made known to the expert at or before the hearing. If of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject, the facts or data need not be admissible in evidence.
"The rationale for this requirement is that expert testimony seeks to assist the trier of fact. An expert opinion that is not reliable is of no assistance to anyone." State v. Kelly, supra, 97 N.J. at 209, 478 A.2d 364 . In criminal cases, the Supreme Court of New Jersey applies the general acceptance or Frye test for determining the scientific reliability of expert testimony:
[W]hile courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.
[ State v. Harvey, supra, 151 N.J. at 169 , 699 A.2d 596 (citing Frye v. United States, 54 App.D.C. 46, 293 F. 1013, 1014 (D.C. Cir. 1923)).]
Although general acceptance does not require complete agreement over the accuracy of the test or infallibility, it encompasses "an extraordinarily high level of proof based on prolonged, controlled, consistent, and validated experience." Rubanick v. Witco Chem. Corp., 125 N.J. 421 , 436, 593 A.2d 733 (1991).
New Jersey courts recognize three ways the proponent of expert testimony can prove general acceptance of a new field or technique, and thereby, its reliability:
(1) by expert testimony as to the general acceptance, among those in its profession, of the premises on which the proffered expert witness based his or her analysis;
(2) by authoritative scientific and legal writings indicating that the scientific community accepts the premises underlying the proffered testimony; and
(3) by judicial opinions that indicate the expert's premises have gained general acceptance. [ Kelly, supra, 97 N.J. at 210, 478 A.2d 364 (citing State v. Cavalio, 88 N.J. 508, 521, 443 A.2d 1020 (1982)).]
The burden to "clearly establish" reliability of the evidence is on the proponent. State v. Harvey, supra, 151 N.J. at 170, 699 A.2d 596 (citing State v. Williams, 252 N.J.Super. 369 , 376, 599 A.2d 960 (Law Div. 1991)).
Proving general acceptance "entails the strict application of the scientific method, which requires an extraordinarily high level of proof based on prolonged, controlled, consistent, and validated experience." Rubanick, supra, 125 N.J. at 436, 593 A.2d 733 . It does not, however, require that the scientific community completely accepts the new technology. State v. Harvey, supra, 151 N.J. at 171, 699 A.2d 596 (citing Biunno, Current N.J. Rules of Evidence, comment 4 to N.J.R.E. 702; State v. Tate, 102 N.J. 64 , 83, 505 A.2d 941 (1986); State v. Johnson, 42 N.J. 146 , 171, 199 A.2d 809 (1964)). New technology is considered to be generally accepted by the relevant scientific community "when it passes from the experimental to the demonstrable stage." Ibid. (citing Windmere, Inc. v. Int'l Ins. Co., 105 N.J. 373 , 378, 522 A.2d 405 (1987)).
Here, the State's burden is to prove that EDR devices, and the interpretation of the data taken from them, is a non-experimental, demonstrable technology that the relevant scientific community widely, but perhaps not unanimously, accepts as reliable.
Expert Testimony
Haight testified that accident reconstruction is the science of determining what happened immediately before and during an accident by examination of the crash scene. He has been involved in over 1000 crash tests since 1985. In about 825 of the tests, Haight was in the car when the crash took place. He is the director of the Collision Safety Institute, a training and research center which operates primarily in the United States, Canada, and Australia. The institute advises doctors, lawyers, police, and anyone else involved in crash analysis.
Haight told the court that GM began putting EDR systems into their vehicles in the early 1990s. By the 1996 model year, most GM vehicles had a modern EDR. By 2000, these devices were able to capture precrash data. Haight described the capturing of precrash data as a "5 second loop," with the airbag computer recording the engine data once per second and storing each recording made during the previous five seconds. If there is a deployment or near-deployment event, the loop is discontinued. Haight told the court that there are about forty-one million cars that have EDR systems that preserve precrash data on the road today.
Haight further testified that he has obtained his knowledge of EDR systems through testing and working with automotive engineers. He began receiving training in EDR technology from GM and Vetronix in 2000. Around that time, Haight began conducting crash tests that compared the rate of vehicle deceleration recorded by the SDM with the rate of deceleration recorded by accelerometers. Haight has testified about the reliability of EDR technology at fourteen Daubert v. Merrell Dow Pharmaceuticals, 43 FM 1311 (9th Cir. 1995), hearings, in states such as California, Florida, New York, Michigan, and Mississippi.
An accelerometer is a device attached by the testers to the vehicle's diagnostic computer, which measures the vehicles rate of deceleration independently of the EDR system.
Regarding the manufacture and testing of the GM SDM, Haight told the court that the manufacturer, Delphi, subjects each SDM to a ten G "shock," and then checks to ensure that the SDM recorded the data accurately. During assembly, GM tests the SDM to ensure that the diagnostic routine that is performed when a vehicle is started is done correctly. In "real world" crash analysis, the most common problem or anomaly with the EDR is incomplete data due to an electrical short. The electrical short does not skew the data; it simply stops the EDR from recording it.
Haight asserted that there are no organizations within the automotive community which question the reliability of EDR devices. He is also of the opinion that the devices are highly reliable.
Besides auto manufacturers, the Insurance Institute for Highway Safety and the National Highway Traffic Safety Administration use EDR data to facilitate auto safety research.
As to the instant case, Haight stated that he analyzed the "raw" data that was recorded by the SDM in defendant's Hummer H2, and did not find any anomalies. He also analyzed the data downloaded by the Vetronix CDR and compared it with photos of the crash scene and what was written in the police report. Haight told the court that he found the photos and police report consistent with the data recorded by the SDM.
"Raw" data is the data that is saved in Hexadecimal on the SDM storage device.
According to Haight, the maximum deceleration of defendant's Hummer during the accident, as measured by its SDM, was only 11.5 miles per hour. This deceleration was not severe enough to cause the Hummer's airbag to deploy. The reason for the non-deployment was the weight discrepancy between the two vehicles; defendant's Hummer weighed 6300 pounds, while the victim's Civic weighed 2300 pounds. In essence, the collision with the Civic did not slow down the Hummer enough for its airbag to deploy. Haight also testified that the rollover of the Hummer, after it struck the victim's vehicle, was "post event," and had no effect on the data recorded by the Hummer's SDM.
Haight also described the process of downloading the SDM data with the Vetronix CDR. Basically, the process involves connecting a diagnostic link connector with the vehicle's computer. Haight told the court that although the persons downloading the data must know what they are doing, the process is simple. He stated that the raw data remained in the SDM, so a defense expert also could download it.
Finally, when questioned by the court, Haight stated that in all the previous cases in which he testified about the reliability of EDR devices, the court accepted his opinion as an expert qualified in his field.
Authoritative Scientific and Legal Writings
In support of the reliability of the CDR data, the State has submitted a downloaded printout from the Vetronix Internet site, which answers questions that are frequently asked about the CDR system, and two published articles. The first article, W.R. Rusty Haight, Automobile Event Data Record (EDR) Technology, 1-2 (2004), discusses the accuracy and precision of the General Motors EDR technology, and how these concepts relate to its reliability. The article states that the SDM is reasonably precise; it samples the vehicle's acceleration every 312 microseconds, and records changes in the vehicle's velocity in ten millisecond intervals. Regarding accuracy, the article states that in tests of "head-on," "broadside," and "rear-ender" ("in-line") collisions, the SDM under-reported the maximum change in vehicle velocity by an average of 1.13 miles per hour with a standard deviation of 1.11. In tests of other types of collisions (those that occur on an angle), the device under-reported the change in velocity by .13 miles per hour. Haight concludes that EDR technology, although evolving, is accurate to a level of precision necessary for it to be reliable for its intended purposes.
The second article, H. Clay Gabler, Use of Event Data Recorder (EDR) Technology for Highway Crash Data Analysis, 1-2 (2004), was prepared by the Transportation Research Board of the National Academies. The article noted that in 2003, the Vetronix CDR download failed in about one-third of attempts; however, sixty-two percent of the failures resulted from lack of permission to download the data. Failure of the EDR due to crash damage or electrical problems accounted for another twenty-five to thirty percent of download failures. The board also recommended several improvements that it felt should be incorporated into EDR systems, including: (1) storing vehicle data from five seconds before until five seconds after the crash; (2) the ability to store three events (instead of two for GM and one for Ford EDR systems), which would capture ninety-four crash events; and (3) expanding the definition of "event" to include roadway departures. No studies of the accuracy or reliability of EDR systems were cited.
In addition to the above articles, an internet publication by Mechanical Forensics Engineering Services, LLC, Wade Bartlett, Vetronix Crash Data Retrieval System, 1-8 (2005) cites independent testing that shows a variation of .3 miles per hour between CDR generated data and Information Society Technologies (IST) accelerometer data, with crash durations matching to a few thousandths of a second. However, in low speed events, particularly where the airbag does not deploy, the SDM may miss the first few miles per hour change in velocity before the system "wakes up."
These articles and the downloaded information from the Vetronix internet site support the notion that the reliability of EDR systems, and more specifically the GM SDM and Vetronix CDR, is accepted within the relevant scientific community. The article by Haight concludes that EDR technology is reliable, and that its reliability is likely to improve over time. The Transportation Research Board report made recommendations as to how to improve EDR systems, and noted some problems with downloading vehicle data using the CDR, but did not mention any reliability problems with existing EDR technology.
In State v. Harvey, supra, 151 N.J. at 173-75, 699 A.2d 596 , the Court held that three published articles, a report from the National Research Counsel, and forty-four presentations, posters, lectures, seminars, and workshops in which forensic scientists discussed issues regarding polymarker-related research, testing, and results, supported a conclusion that the polymarker DNA test was accepted by the relevant scientific community. While the Harvey Court has "recognized a correlation between the number of published articles and the general acceptance of a subject . . . it has never required a specific number of published articles to satisfy the test of general acceptance. Rather [its] focus always has been whether existing literature reveals a consensus of acceptance regarding a technology." Id.
In this case, although the literature on the reliability of the data downloaded from GM's SDM, or EDR devices in general, is relatively sparse, the consensus is that the data is reliable. All of the test results that have been reported indicated that the difference in the numbers on vehicle speed and change in velocity taken from the CDS are not significantly different from the numbers recorded by other measuring devices. Most of that difference has been in low-speed "non-deployment" events, which is definitely not the type of event that happened here.
An important factor in resolving the scientific reliability issue is the recognition of the "black box" by the National Highway Traffic Safety Administration (NHTSA), a federal agency which was created by an act of Congress 49 U.S.C. 30101 to:
reduce traffic accidents and deaths and injuries resulting from traffic accidents. Therefore it is necessary —
(1) to prescribe motor vehicle safety standards for motor vehicles and motor vehicle equipment in interstate commerce; and
(2) to carry out needed safety research and development.
When NHTSA mandated airbags, it also required sensing methods to determine when to deploy them. Initially, these were mechanical systems which later developed into electronic systems. The EDR was originally a component of the Air Bag Control Module (ABCM). Its purpose was to monitor the efficiency of the airbag system. The current EDR evolved so that the data recorded is regularly analyzed and relied on by the NHTSA, government agencies, and manufacturers. NHTSA urges voluntary installation and has pending rules to require more data from manufacturers. Event Data Recorder Applications for Highway Safety, 63 Fed.Reg. 60270, Nov. 9, 1998, and 64 Fed.Reg. 29616, June 2, 1999. In sum, when examined in light of judicial opinion and Haight's testimony at the Frye hearing, industry publications regarding the General Motors SDM Vetronix CDR indicate that the traffic collision reconstruction community finds the technology to be reliable. See In re Commitment of R.S 173 N.J. 134 , 136, 801 A.2d 219 (2002) (determination of reliability of actuarial risk assessment instruments to estimate future dangerousness of sex offenders was founded on "validation studies, cross-validation studies, reliability studies, correlation coefficients, and clinically-derived factors").
Judicial Opinions
No New Jersey court has addressed the issue of whether the data from EDR devices is reliable enough to be admitted into evidence at a criminal trial. However, at least seven courts in other states have held that black box data is scientifically reliable.
In Bachman v. GMC, 776 N.E.2d 262 (Ill.App. 2002), a product liability case, the court determined that, with respect to the Frye standard, the process of recording and downloading SDM data was not a "novel technique or method." Noting that crash sensors such as the SDM had been in production in automobiles for over a decade, and that the microprocessors that run them and record their data also run everyday appliances, such as computers and televisions, the court stated that "[i]f the scientific evidence is not `novel,' then the Frye admissibility standard has been satisfied." Id. at 262. The court concluded that the trial court had not abused its discretion by finding that the process of recording and downloading SDM data was sufficiently established to have gained general acceptance in the relevant scientific community because the process was not the sort of method "new to science that undeservedly creates a perception of certainty when the basis for the evidence or opinion is actually invalid." Id. at 281 ( quoting Donaldson v. Cent III. Pub. Serv. Co., 199 Ill.2d 63, 262 Ill.Dec. 854, 767 N.E.2d 314 , 324 (2002)). In Matos v. State, 899 So.2d 403 , 405 (Fla.App. 2005), two teenage girls were killed when the defendant's car struck their vehicle in a residential neighborhood. A "black box" computer which operated the airbag in the defendant's car was recovered after the crash. Ibid. The EDR showed a speed of 114 miles per hour for the defendant's car four seconds prior to the crash and a speed of 103 miles per hour within one second of the crash. Ibid. Evidence showed that the defendant's airbag was working properly at the time of the accident. Ibid. The defendant argued that the recorded evidence did not meet the Frye test for admissibility and therefore was admitted in error. Ibid.
The court in Bachman cited the American Heritage Dictionary of the English Language 898 (1975) definition of the word "novel" as "strikingly new, unusual, or different." Bachman, 332 Ill.App.3d 760, 267 Ill.Dec. 125, 776 N.E.2d 262 , 281 (2002).
In order to show the reliability of the black box data, the State of Florida called two experts to testify at the Frye hearing. Ibid. The first expert, an accident reconstructionist, testified that information from the SDM is generally accepted in the accident crash investigation community, in the insurance field, and in medical research and biomechanics. Ibid. He also testified that SDM data was used by the NHTSA. Ibid. The state's second expert, an industrial engineer, testified that data collected by SDMs are generally accepted within the fields of automobile safety, accident reconstruction, and automotive design. Id. at 406. He testified that the NHTSA has had their crash teams using the data since around 1995. Ibid. He further testified that the SDM is extremely accurate because it is a digital system, and that the data collected from the SDM reflects the electronic system of the car. Ibid. The expert did acknowledge that he utilizes other crash information to verify the accuracy of the data. Ibid.
The state also introduced two industry papers. The first was a report by the Society of Automotive Engineers (SAE) entitled "Accuracy of Pre-Crash Speed Captured by Event Data Recorders," that concluded that the EDR data overestimated vehicle speeds by one mile per hour at low speeds and by 2.5 miles per hour at high speeds. Ibid. The second report, authored by the National Highway Transportation Safety Board and GM engineers, cited a study on real life crashes which indicated that the SDM recording of vehicle speed was accurate within plus or minus four percent. Ibid.
The court concluded that the process of recording and downloading SDM data was not a novel technique or method, and that when used as a tool of automotive accident reconstruction, the SDM data was generally accepted in the relevant scientific community, warranting its introduction. Id. at 407.
In addition to the two cases described above, three New York State trial courts have found that a Frye hearing on the admissibility of General Motors SDM recording and download procedure was not necessary because these technologies were well established and accepted within the relevant scientific community. People v. Christmann, 3 Misc.3d 309, 776 N.Y.S.2d 437 (N.Y.Just. Ct., 2004). This court agrees that SDMs were generally accepted in the field after Haight testified that they were "extremely reliable."
A constitutional rights issue is raised as to the propriety of the use of the property owned by defendant (the black box in his car) against him. Suffice it to say that defendant's right against self-incrimination is not violated where the evidence is not testimonial. Rawlings v. Police Dep't of Jersey City, 133 N.J. 182 , 194, 627 A.2d 602 (1993) (citing Schmerber v. California, 384 U.S. 757 , 761, 86 S.Ct 1826, 1830, 16 L.Ed.2d 908, 914 (1966)). The use of DNA, blood, and breath samples have all been permitted as constitutionally acceptable. Schmerber, supra, 384 US. at 765, 86 S.Ct. at 1832, 16 L.Ed.2d at 916 ; State v. Stever, 107 N.J. 543 , 558, 527 A.2d 408 (1987); State v. King, 44 N.J. 346 , 357, 209 A.2d 110 (1965); State v. Macuk, 57 N.J. 1 , 14, 268 A.2d 1 (1970). Defendant's property was properly seized under a search warrant based upon probable cause, and defendant is free to retain an expert to challenge the testimony of the State's witness's conclusions.
The Court concludes that the scientific reliability of event data recorder evidence was generally accepted within the automotive and accident reconstruction community, and thus met the Frye standards for admissibility.