Deciphering the Sensory Data Module
February 13, 2008
Okay, so what is an Sensory Data Module? If you have to ask, you need to know. Whether your case load consists of personal injury or DWI cases, you’ll undoubtedly deal with auto accidents. If one of the vehicles involved is a 1995 or later and was manufactured by General Motors, it contains an SDM (or more properly SDM-R).
A SDM-R or Sensory Data Module Recorder, is an electronic data recorder which has been fitted to most GM vehicles since 1995. At its web site, General Motors sets out that the purpose of these devices is to analyze “real world collisions for the purpose of assessing, and improving, crash performance.” Tellingly, they also note that “[e]xaminations occur for a variety of reasons including customer inquiries and claims, requests by government officials, dealers or others, and for GM's own crash performance research.”
While GM declares that the SDM records speed, engine RPM, brake and throttle application data, our research indicates somewhat more. The SDM is a computer which is linked electronically to the vehicle’s airbag sensors. When the sensor detects abrupt vehicle deceleration, the SDM “wakes up” and begins to record data to a EEPROM or erasable programmable read‑only memory chip. Initial versions of the module recorded the time between SDM initiation and airbag detonation, the maximum delta V (the change in velocity), the delta V in a frontal collision, the time between the moment the vehicles impacted and the moment of maximum delta V. Additionally, in 1999, the SDM was beefed-up to include the ability to record vehicle speed, engine speed, brake switch and percent of the throttle status. Erasure incidents include a deployment or near deployment such as bumping the vehicle into a curb, hitting a pothole, or suddenly engaging the brakes causing a faster deceleration than what occurred in an accident. Further, the data will be erased after the car ignition is turned on 250 times. Even if the vehicle pre-dates 1999 by as much as one year, the extended features may nonetheless be included by virtue of the fact that GM has a reputation for quietly introducing electronic enhancements prior to them being announced.
When handling a DWI accident, a frequent issue will be the cause of the accident as New York law has been quite clear for over three decades that that is still the rule. Simply stated, intoxication alone, whether by alcohol or drugs, is an insufficient basis upon which to secure a conviction for Criminally Negligent Homicide (see, People v. Bast (1967) 19 N.Y.2d 813, 280 N.Y.S.2d 149; see also, Re: Johnston (1990) 75 N.Y.2d 403, 554 N.Y.S.2d 88) and presumably vehicular assault as causation is still a necessary element (see, 7 NY DWI Bulletin 25). This is where the SDM comes into play. While in the past we have relied upon techniques such as Crush Analysis (see, 7 NY DWI Bulletin 25), the presence of an SDM significantly alters the equation inasmuch as a presumably accurate record of vehicle speed and some pre-accident operations will be preserved.
Two years ago, our office handled what we believe was the first matter in New York to involve the presence of a SDM. Just how the accident occurred was the subject of true debate as traditional accident reconstruction was not conducted until almost two months after the collision and well after debris and telltale signs had been washed away. Further, intoxication was very much an issue as the blood was drawn under circumstances which certainly mandated suppression. As luck would have it, the SDM was destroyed the very day the defendant was indicted. Fortunately, for our client, the confluence of these factors led to a very favorable plea offer, although this meant that the SDM issues, including reliability and destruction, would never see a decision.
In the two years since our matter was litigated, there has been some activity on the SDM front. Given the overwhelming importance of these ubiquitous little devices to what we do on a daily basis, we decided to review every available case.
In Bachman v. General Motors Corporation, 332 Ill. App. 3rd 760 (4th Dist 2003), the plaintiff brought suit against General Motors Corporation claiming that the airbag in her 1996 Chevrolet Cavalier detonated prematurely causing the collision in which she was injured. Following a defense verdict, the plaintiff appealed contending, inter alia, that the trial court erred by allowing evidence regarding data downloaded from the Cavalier's sensing and diagnostic module and admitting related opinion testimony. In essence, plaintiff, in this civil case, was calling for a Frey hearing (see, Frye v. United States, 293 F. 1013 [D.C.Cir.1923]; and see, People v. Leone, 25 N.Y.2d 511, 307 N.Y.S.2d 430, 255 N.E.2d 696 ) on the reliability of the SDM-R. In October of 2000, the hearing was held and since the developer of the SDM (or SDM-R) testified, the results were quite interesting and deserve an extended examination.
Douglas Nunan, a senior project engineer in Delphi's restraint systems electronics group, testified that between August 1989 and May 1996, he was responsible for designing and developing SDMs, including the type of SDM installed in Danielle Bachman's Cavalier. He testified that the SDM, which is controlled by a microprocessor, has multiple functions: (1) it determines if a severe enough impact has occurred to warrant deployment of the air bag; (2) it monitors the air bag's components; and (3) it permanently records information. He further testified that the SDM contains software that analyzes the longitudinal deceleration of a vehicle to determine whether a deployment event has occurred “based on testing that was done previously to determine what events would require protection by an air bag.” When the SDM senses an event, either a deployment event or, interestingly, an event that is not severe enough to require an air bag — that is, a near‑deployment event — that information is recorded to the microprocessor's electrically erasable programmable read‑only memory (EEPROM). When the air bag is deployed, the SDM records the event as a “Code 51.” Nunan explained that when an event is written to the memory, “you have a record to go back and look to see, did the result that came out match what you should have had.” Although certain diagnostic codes stored in the EEPROM could be erased, the SDM installed in the Cavalier “was specifically designed to prevent” air‑bag deployment data from being altered or erased. Additionally, Nunan opined that the type of SDM installed in the plaintiff’s Cavalier was accurate and reliable.
Nunan explained that the SDM data is stored in hexadecimal format. Therefore it must be converted to a decimal format before it can be analyzed. (A hexadecimal system is a numbering system that has 16 characters; that system employs the letters A through F, in addition to the numbers zero through nine, which are used in the decimal system.) While a competent engineer could convert and read the data using a basic calculator and General Motors documents, Nunan explained how in mid-2000, a publicly available crash data retrieval tool (Vetronix equipment) was updated with new software to allow anyone with a Windows‑based computer to download SDM data in an easy‑to‑read format.
One of the more common questions concerning SDM usage is whether the loss of power on the vehicle will affect the existence or accuracy of the data previously retrieved. In this regard, Nunan testified that a power loss during a crash would not affect data previously recorded. An external power source can be used to retrieve data from the SDM. According to Nunan, such a procedure is normal and does not affect the accuracy of the SDM's data. Nunan further stated that the newly updated Vetronix equipment could be used to retrieve data from an SDM and that data would be identical to the data originally retrieved. He acknowledged that the SDM's performance specifications were not public knowledge. He also acknowledged that at the time of the collision, crash data recorders in vehicles were not standardized.
John Sprague, a Delphi systems engineer who worked at General Motors in Warren, Michigan, also testified at the Frey hearing. He testified that from 1993 through 1996, he was a sensing application engineer for several vehicles. He conducted developmental testing of the SDM and investigated actual field deployments. The SDM, on the 1996 Cavalier, was designed to sense longitudinal deceleration of the vehicle. He too acknowledged that the 1996 Cavalier SDM's product definition documents were "generally treated as confidential" by General Motors. Sprague opined that neither the defect nor the recall had anything to do with the SDM's recording function, the retrieval of data from the SDM, or the reliability of its recorded data. General Motors submitted information regarding the investigation to the NHTSA, which must approve the determination and proposed correction of an automobile defect. According to Sprague, no one from NHTSA ever questioned the reliability or accuracy of the data being downloaded from the 1996 J‑cars. Based upon his developmental testing of the SDM and his investigation of the inadvertent‑deployment events, Sprague opined that the SDM's recording function was both reliable and accurate.
Both GM witnesses testified that the 1996 Cavalier had been recalled due to early airbag deployment issues. Both testified, however, that this problem and the subsequent recall had nothing to do with the accuracy of the data which was subsequently retrieved.
Finally, the Bachman court also considered several affidavits and documents submitted by the parties. Donald Floyd, a former supervisor of diagnostic software activities at General Motors, averred that he was responsible for releasing data to Vetronix for its use in developing a crash retrieval system for General Motors’ cars. He set out that the Vetronix equipment allows anyone with a computer equipped with Windows 95 or 98 to download data recorded by SDMs on General Motors’ vehicles. Floyd set out that he participated in testing with Vetronix, the Insurance Institute for Highway Safety, and Canadian law enforcement, in which the accuracy of downloaded SDM data was confirmed.
Daniel Faust, a General Motors staff development engineer, averred that the NHTSA and the National Transportation Safety Board (NTSB) have requested other downloaded SDM collision data from General Motors.
Keith Schultz, a General Motors senior staff engineer, averred that other government and law enforcement agencies also request and rely on SDM collision data downloaded by General Motors’ engineers. Schultz and NHTSA representatives co-authored an article entitled “Recording Automotive Crash Event Data,” in which they concluded that the loss of electrical power during a crash did not affect the reliability or accuracy of data retrieved from the SDM and suggested traffic safety uses for the SDM data.
Needless to say, at the conclusion of the hearing, the Michigan trial court concluded that the SDM (or SDM-R) met the Frey standard for reliability and scientific acceptance.
Early this year, the SDM (or SDM-R) made its entry into New York case law. It did so by means of a charge for violating of §1180 (d) of the Vehicle and Traffic Law and Failure to Exercise Due Care in violation of Vehicle and Traffic Law § 1146. In People v. Christmann, 3 Misc.3rd 309, 776 N.Y.S.2d 437 (Newark Village Court, January 16, 2004) Newark Village Justice V. Bruce Chambers explored the issues present in Bachman, with an interesting twist. Christmann, was a fatal accident in which a pedestrian was struck. Apparently, at issue was the speed the defendant was proceeding immediately prior to the crash. Accordingly, Trooper Robert J. Frost, who has been assigned to conduct accident reconstruction for the New York State Police, performed “traditional” accident reconstruction such as measurement of skid marks, extensive photographing of the scene, determination of the relative position of the vehicle fragments and personal property, and determination of the point of impact. Of importance to our topic, Trooper Frost used computer equipment in his police car to download information from the sensing diagnostic module (SDM) located in the defendant's vehicle. He conducted this procedure without seeking or obtaining the permission of the defendant. Thereafter, the People sought to utilize the data retrieved from the SDM at a trial held on the uniform traffic citations.
Christmann, therefore presents at least two issues. First, was the data from the SDM scientifically reliable and accepted as required under the so-called Frey/Leone standard (see, Frye v. United States, 293 F. 1013 [D.C.Cir.1923]; People v. Leone, 25 N.Y.2d 511, 307 N.Y.S.2d 430, 255 N.E.2d 696 ) and if so, what type of foundation need be established? Secondly, did the warrantless retrieval of the SDM data from the defendant’s vehicle violate the defendant’s Constitutional protections against warrantless search and seizure as protected under the 4th Amendment to the United States Constitution and Article I § 6 of the New York State Constitution?
Concerning the first issue, the court, following Bachman, held:
[t]he admissibility of evidence of the data recorded on a SDM has been received into evidence as ‘generally accepted as reliable and accurate by the automobile industry and the [National Highway and Traffic Safety Administration].’ See, Bachman v General Motors Corp. (332 Ill App 3d 760, 768, 766 NE2d 262, 272 [4th Dist 2002]), which held that such evidence was admissible under the standards of Frye v United States (293 F 1013 ).
The other critical issue that remained was the search and seizure issue. As noted above, as part of his investigation, the Trooper retrieved the data from the defendant’s SDM. To do this, he conducted this procedure without seeking or obtaining the permission of the defendant. To do so, Trooper Frost asserted control over the vehicle, directing it not be moved until his investigation was completed. After accomplishing the tests, the vehicle was returned to the defendant. In the words of the court, Trooper Frost thus "impounded" the vehicle even if for a short period of time. While the vehicle was in his possession, he also operated it to confirm that the brakes were working properly and utilized an accelerometer in order to measure the braking capability of the car.
The question is the obvious, did Trooper Frost require a warrant to assert control over the vehicle and did his retrieval of the data similarly require a warrant?
Holding that a warrant was not required, the court initially turned to Vehicle and Traffic Law §603(1) and Vehicle and Traffic Law §603-a. Enacted in 1993, Vehicle and Traffic Law §603(1) provides, in pertinent part, that “[e]very police . . . officer to whom an accident resulting in injury to a person shall have been reported . . . shall immediately investigate the facts . . . and report the matter to the commissioner [of Motor Vehicles].” Enacted in 2001, §603-a provides that any motor vehicle accident reported or discovered by a police officer and the accident resulted in serious physical injury or death shall be investigated by the officer. The statute further provides that “[s]uch investigation shall be conducted for the purposes of making a determination of the following: the facts and circumstances of the accident; . . . the contributing factor or factors; whether it can be determined if a violation or violations of this chapter occurred; . . . and, the cause of such accident" (Vehicle and Traffic Law § 603‑a ).
In People v. Quackenbush, 88 N.Y.2d 534, 647 N.Y.S.2d 150  the Court of Appeals had occasion to consider the Constitutionality of Vehicle and Traffic Law §603(2). This statute provided:
2. In addition to the requirements of subdivision one of this section, every police officer or judicial officer to whom an accident shall have been reported involving a commercial vehicle as defined in either subdivision four of section five hundred one-a or subdivision one of section five hundred nine-p of this chapter shall immediately investigate the facts, or cause the same to be investigated and report the matter to the commissioner forthwith, provided that the report of the accident is made to the police officer or judicial officer within five days after such accident, whenever such accident has resulted in (i) a vehicle being towed from the accident scene as the result of incurring disabling damage, (ii) a fatality, or (iii) any individual being transported to a medical facility to receive treatment as the result of physical injury sustained in the accident.
In Quackenbush (supra), the defendant was involved in a fatal pedestrian accident involving a bicyclist. The police impounded his car and two days later conducted a safety inspection of the equipment of the vehicle, including brakes. The brakes were found to be deficient and defendant was charged with the misdemeanor of Operating with Defective Brakes in violation of Vehicle and Traffic Law § 375(1). Thereafter, the defendant moved to suppress the results of the inspection upon the same grounds that the vehicle had been subject to an unconstitutional seizure. Successful before the East Hampton Town Court, the Appellate Term of the Supreme Court in the Second Judicial Department reversed. Affirming the Order of the Appellate Term, the Court of Appeals concluded that “the police possessed the authority to impound the vehicle in order to comply with the investigatory and reporting duties imposed by Vehicle and Traffic Law § 603” (Quackenbush, at 537). The court stated that because a vehicle's safety equipment was subject to “extensive government regulations,” including mandatory annual inspection, that a safety inspection after a fatal accident “did not offend the constitutional prohibitions against unreasonable searches and seizures” (Id.).
Having once determined that the impoundment procedure was satisfactory, the court then considered the process of the subsequent search. Clearly, there was no exigency involving mobility of the vehicle as it was under police control. Holding that only unreasonable searches violative of expectations of privacy were proscribed and finding that the search there was justified at the inception and limited in scope, the court upheld the search and denied the motion to suppress. The court found only a diminished expectation of privacy in the mechanical areas of the vehicle and further found that that expectation must yield to the overwhelming state interest in investigating fatal accidents (People v. Quackenbush, supra. at 539). The court also found that the search conducted of the safety equipment of the truck in question was of an administrative nature, rather than an attempt to gather information to form the basis of a criminal prosecution:
Because of this extensive regulation of vehicular safety equipment, there is only a diminished expectation of privacy in the mechanical areas of a vehicle. [Footnote omitted]. When a fatal accident involving an automobile has occurred, that minimal privacy expectation necessarily yields to the State's legitimate public safety interests in determining all of the circumstances surrounding the death and the cause of the accident (People v Ingle, 36 NY2d 413, supra) and in ensuring that the vehicle is not returned to the roadway in an unsafe condition. As one jurisdiction has noted in reaching a similar conclusion, "[s]ociety places great importance on learning all the circumstances of any motor vehicle accident resulting in death", and would not "recognize as objectively reasonable an expectation of privacy in the braking mechanism of a motor vehicle that had come into police possession following the death of a motorist on the highway" (Commonwealth v Mamacos, 409 Mass 635, 640, 568 NE2d 1139, 1142).
Returning to Christmann, the court found “[i]n the case at bar the intrusion sought to be prohibited is significantly less. In Quackenbush, the whole vehicle was seized, taken from the scene, held for over two days and partially taken apart. Here, the vehicle wasn't moved, only one door was opened, a sampling taken and the car immediately returned to the defendant.”
In similar fashion, the court found exigent circumstances to be present for the “search.”
We now turn to the concept of exigency. While a pure "automobile exception" does not apply since the officer had no probable cause to believe that evidence of a crime was contained within the car, courts have upheld warrantless searches of automobiles based upon exigency. (Carroll v United States, 267 US 132 .) Such an exception now has been mostly replaced by findings that automobiles contain a diminished expectation of privacy. (Cardwell v Lewis, 417 US 583 ; Robbins v California, 453 US 420 .) Here, however, real exigency exists. Evidence regarding the pre-accident conditions within defendant's automobile could easily be destroyed, either purposely or accidently, if the automobile was moved from the scene under its own power.
Accordingly, the court held:
While Quackenbush would appear to allow impoundment of the vehicle and subsequent inspection pursuant to the authority of Vehicle and Traffic Law § 603 (now enhanced by Vehicle and Traffic Law § 603‑a), I conclude that the immediate download of information from the defendant's SDM is permitted and required by Vehicle and Traffic Law § 603‑a and is not violative of defendant's rights to be free from unreasonable searches pursuant to the United States or New York Constitution.
While Christmann may have been decided correctly, we fear that the court missed a step. The crux of the motion was not to suppress evidence of pervasively regulated areas such as breaks, tires and steering, but to suppress the retrieval of data. Was the data constitutionally protected? While the court was quite correct that the exigency requirement upon which the automobile exception created in Carroll v. United States, 267 U.S. 132 , has been largely superceded by an analysis founded upon “a diminished expectation of privacy” (see, South Dakota v. Opperman, 428 U.S. 364, 96 S.Ct. 3092, 49 L.Ed.2d 1000), the diminished expectation runs to what may be seen by others due to the visible nature of an automobile. Can the same be said of computer data that requires a specialized program for retrieval? Does this mean that a lap top computer that is on the front seat of the car may be similarly accessed? Isn’t the data situation analogous to a container contained in a vehicle? The answer to this is yes, however, in California v. Acevedo, 500 U.S. 565, 111 S.Ct. 1982, 114 L.Ed.2d 619 (1991), the court specifically ruled that the police can enter closed containers found in a vehicle if they have probable cause to believe that the container contains evidence of a crime. It would be our humble opinion that the SDM data at the scene of a wreck is analogous to the closed container. The accident provides probable cause to believe that the computer may contain data and it may be accessed accordingly.
The latest courtroom appearance for the SDM occurred in Rochester, New York. In People v. Hopkins, the evidence before the Grand Jury was that a Cadillac struck the rear-end of a Dodge Neon and the collision resulted in the death of Lindsay Kyle, the driver of the Neon. The evidence apparently demonstrated that, shortly before the collision, a witness driving on Exchange Boulevard near the Interstate Route 490 overpass, observed the Cadillac pass his vehicle on the right going at least double his speed of 30 or 35 miles per hour. The Rochester Police Department’s reconstruction of the accident, based upon data collected by the New York State Police and from the air bag module in the 2004 Cadillac, demonstrated that the estimated speed at which the defendant’s vehicle was traveling at the time of impact was 65-70 miles per hour and that the brake switch was on approximately two and a half to three seconds prior to impact. The RPD reconstruction specialist testified additionally that the data from the air bag module showed that five seconds before impact, the rate of speed of the Cadillac was 104 miles per hour and that three to four seconds before impact, the vehicle was traveling at 106 miles per hour. He also testified that the road was wet on the night the collision took place.
Charged with Depraved Indifference Murder (Penal Law § 125.25 ), speed was an invaluable component of the prosecution’s case. Given the fact that the SDM comprised a critical portion of that proof, the defendant requested a Frey hearing on the reliability of the module. Denying the hearing, the Monroe County Court (Geraci, J.) turned to the standard for scientific tests as recently annunciated by the Court of Appeals in People v. Wesley, 83 N.Y.2d 417:
While foundation concerns itself with the adequacy of the specific procedures used to generate the particular evidence to be admitted, the test pursuant to Frye v United States (293 F 1013) poses the more elemental question of whether the accepted techniques, when properly performed, generate results accepted as reliable within the scientific community generally. *** The long-recognized rule of Frye v United States (supra) is that expert testimony based on scientific principles or procedures is admissible but only after a principle or procedure has ‘gained general acceptance’ in its specified field. In Frye (supra, at 1014) the court stated: ‘Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while 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’ (emphasis supplied). *** This Court has noted that the particular procedure need not be ‘unanimously indorsed’ by the scientific community but must be ‘generally accepted as reliable’ (see, People v Middleton, 54 NY2d 42, 42, 49, supra).
The court then turned to Bachman, Christmann and several scientific articles that were submitted by the parties on the motion and concluded:
The key to admissibility of novel or scientific evidence is general acceptance in the scientific community as reliable. According to one noted treatise, reliability may be established in at least three ways: (1) “general acceptance may be so notorious that the court may take judicial notice of it” (2) “acceptance may be established by reference to legal writings and judicial opinions” (3) if neither of the above can be resorted to, the trial judge “may conduct a hearing at which the proponent may establish admissibility by offering evidence of acceptance, including the expert’s own testimony” (Prince, Richardson on Evidence § 7-311 [Farrell 11th ed]; cites omitted).
In this case, the court is persuaded, based upon its review of the cases and other supporting documentation submitted by the People, and in the absence of any contrary or contradictory evidence, that the SDM module technology has been generally accepted as reliable in the relevant scientific community. A Frye hearing is, therefore, unnecessary to determine the admissibility of evidence with respect thereto at trial. The defendant’s request for such hearing is denied.
« Go Back