Wednesday, August 04, 2010

Daubert and Cell-Site Data

As I explained in an earlier post, the U.S. Supreme Court’s decision in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993) established the test federal (and many state) judges currently use to determine the admissibility of expert testimony in trials and other proceedings.

The admissibility of expert testimony in federal cases is governed by Rule 702 of the Federal Rules of Evidence, which provides as follows:

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, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.

As I noted in my earlier post, the Daubert Court held that when one side wants to present expert testimony, the judge must determine “whether the expert is proposing to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine a fact in issue.” Daubert v. Merrell Dow, supra. The Court explained that this “entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.” Daubert v. Merrell Dow, supra.

The Daubert Court noted that “many factors will bear on this inquiry” and said it would “not presume to set out a definitive checklist or test” to be used in this assessment. Daubert v. Merrell Dow, supra. The Court also articulated four criteria courts should consider in determining the validity of an expert’s methodology: whether the theory has been or can be tested or falsified; whether the theory or technique has been subject to peer review and publication; whether there are known or potential rates of error with regard to specific techniques; and whether the theory or approach has general acceptance. U.S. v. Benford, 2010 WL 2346305 (U.S. District Court for the Northern District of Indiana 2010).

That brings us to the Benford case: To understand how Nichelle Benford wound up using Daubert in an attempt to block the use of certain expert testimony, we have to start with what happened in northern Indiana on February 13, 2003 and thereafter:

[A]t about 5:21 a .m., a witness noticed that Car's Towing in Schererville . . . was on fire. Investigators questioned Brian Booker, who was suspected of starting the fire. Booker told the investigators that he was with Benford during the early morning hours of February 13th.

The Government contends that local police used a State of Indiana subpoena to obtain the cell phone records belonging to Benford and Booker. The information obtained includes `cell-site data’; essentially, this data reveals the general vicinity of where a cell phone was located at a particular point in time by identifying which cell tower communicated with the cell phone while it was turned on. The Government contends that Benford's cell-site information shows Benford's cell phone communicated with cell phone towers in Chicago around the time of the fire, thus refuting Booker's alibi.

Benford was subpoenaed to testify before a grand jury regarding the Government's case against Booker. . . . Benford testified before the grand jury that she was indeed with Booker around the time when the fire purportedly started.

U.S. v. Benford, supra. Benford was charged with “making false statements before a grand jury” and the prosecution informed the court that it intended to use the cell-site data, “as interpreted by” an expert witness, as part of its case" against her. U.S. v. Benford, supra. The witness was Anthony Imrisek:

The Government seeks to call Imrisek as an expert witness to testify regarding a technique whereby phone calls allegedly made from defendant's cell phone around the time she told a grand jury she was with her boyfriend Brian Booker (thus providing Booker with an alibi for a fire he is suspected of having started) can be used to determine the approximate location of defendant when she allegedly made the calls.

U.S. v. Benford, supra. Benford filed a motion to exclude Imrisek’s testimony as not satisfying the requirements of Rule 702 of the Federal Rules of Evidence as construed by the Supreme Court in Daubert. U.S. v. Benford, supra. In ruling on the motion, the judge who has the case was required to apply the Daubert standards (and factors) outlined above. To establish a factual foundation for that analysis, the judge held a hearing on the Daubert issue, at which Imrisek testified “regarding his background and the methodology used in arriving at his conclusions.” U.S. v. Benford, supra.

After considering the evidence presented at the hearing and the arguments made by both sides (none of which are included in the opinion, so I’m inferring that Benford challenged the relevance of the testimony and/or the reliability of Imrisek’s testimony and/or the reliability of the methodology he used in arriving at the conclusions about which he would testify), the judge issued his ruling. U.S. v. Benford, supra.

The district court judge found, first, that Imrisek’s testimony was relevant to the issue in the case:

[T]he relevance of Imrisek's anticipated testimony is clear. Benford is charged with lying to a grand jury about her whereabouts on February 13, 2003, to provide an alibi for Booker; to the extent that her cell phone data suggests that she was not where she told the grand jury she was, the data and expert testimony related to the data are relevant.

U.S. v. Benford, supra. He then found that Imrisek’s “anticipated expert testimony” was reliable:

A district court has great latitude in determining not only how to measure the reliability of the proposed expert testimony but also whether the testimony is, in fact, reliable. U.S. v. Pansier, 576 F.3d 726 (U.S. Court of Appeals for the 7th Circuit 2009). The Seventh Circuit has advised that `[t]o determine reliability, the court should consider the proposed expert's full range of experience and training, as well as the methodology used to arrive a particular conclusion.’ Id.

At the Daubert hearing, Imrisek testified that he has a Bachelor's degree in electrical engineering. . . . [and] is employed as a radio frequency (`RF’) engineer for Ericsson . . . assisting with management of its cell phone network. Imrisek has held the position of RF engineer since 1997. As part of his job duties, Imrisek has acquired personal knowledge of the cell towers in the Chicago area, and the coverage areas that relate to those towers. Imrisek's job duties at Ericsson include determining cell-site coverage. These facts indicate that Imrisek's full range of experience and training make Imrisek qualified to testify as to Benford's cell-site data and that his testimony will be reliable.

U.S. v. Benford, supra. The judge also found that Imrisek’s methodology was reliable, based on the application of the four Daubert criteria outlined above:

[At the Daubert hearing,] Imrisek testified. . . . [that when] a cell phone makes a call, it exchanges information with a `cell site’ . . . which is often set on a building or water tower. A cell site has a `coverage area,’ which is a radius of space in which that cell site could communicate with a cell phone. Ericsson maintains records regarding the time a call is originated and terminated, and the cell-site tower associated with the beginning and end of a call. In short, Ericsson maintains `cell-site data,’ which shows which cell phone tower handled a particular call.

Using Benford's cell phone call records for February 13, 2003, Imrisek employed a `prediction tool,’ which generated a map of the coverage area for each of the cell sites associated with each of Benford's cell phone calls. Imrisek created a map with a number of circles representing the approximate locations where Benford's cell phone was used. Imrisek testified that in order for a cell phone to connect with a cell site, it would have to be within the cell site's coverage area. Imrisek often creates these coverage maps for business purposes, including to assist Ericsson's sales force.

Benford's counsel argued that Imrisek's methodology has not been subjected to peer review and that Imrisek did not know the potential rates of error associated with his methodology. However, these considerations are simply a few of many factors that the court is to consider. Imrisek testified that there is no dispute in his technical community as to the methods he used to analyze Benford's cell-site data or to create the coverage area maps using the prediction tool. Imrisek's reliance on Ericsson's data and reports served as a reliable basis for his testimony because these materials are facts or data of a type reasonably relied upon by experts in the field of cell phone and cell tower technology. . . . [T]he court has satisfied itself that Imrisek's analysis yielded facts and data sufficient to support his proposed testimony.

U.S. v. Benford, supra. Finally, the judge noted that the methods Imrisek used

appear subject to testing and falsification. . . . Though Benford's counsel emphasized that the prediction tool used by Imrisek merely predicted coverage areas, Imrisek testified that Ericsson constantly conducts tests by using cell phones and tracking their connections to cell sites, which keeps their predictions as accurate and up-to-date as possible. Benford's counsel also highlighted the fact that Imrisek primarily creates maps and analyzes data for business purposes. However, `the question is not whether [Imrisek] was a professional witness; it is whether he was an expert’ on the intricacies of cell sites and the associated data. . . . The court finds that he is.

U.S. v. Benford, supra (quoting U.S. v. Parra, 402 F.3d 752 (U.S. Court of Appeals for the 7th Circuit 2005)). The judge therefore denied Benford’s motion to exclude Imrisek’s testimony at trial. U.S. v. Benford, supra.


Anonymous said...

The records should be like polygraphs. Inadmissible. There are various ways it's collected depending on networks, who owns the tower, internal software and then there are more ways the data is translated to paper reports depending on software used. A young man was recently convicted in McHenry County IL based on cell site record showing an outgoing call from a nearby town 1 hour after the crime occured. No physical evidence was presented. The mother was overheard saying the bill detail showed it was an incoming call from that town.

Anonymous said...
This comment has been removed by a blog administrator.
Greg Smith said...

Susan, I have only read what is posted at your blog regarding this case. However, it isn't clear what data the expert (Imrisek) actually used. Creating prediction maps provides useful guidance as to where the network anticipates a user may use coverage. The maps can include a range of details reflecting the Mast's (so to speak) location, configuration and arrangements at the transceivers, transmission technology, and radio survey data along certain routes (not every route) being fed back that collectively may be used to generate a map.

I advocate the use of best server plots/density maps and single cell predication plots/density maps. I recall a dispute with a prosecution expert who suggested they are unreliable. The basis of his generalised statement in his report was put to the test as to how, if the pros expert had assessed each matter in the case on a case-by-case basis, why use general copy and paste arguments in his reports as a reason for not getting the necessary maps. The expert argued that the murder was conducted in a first flat and the maps would only show coverage at ground level. It was pointed out that the pivotal part of the prosecution mobile case against the defendant he was aware was a burnt out car and this was not found in a first floor flat but on the ground and that without his cell site analysis opinion on coverage in the area of the burnt out car there was nothing to link the burnt out car to the mobile evidence and thus to the defendant. The expert he reassessed the value of his first opinion after that. There was much more but that was a little bit connected to the maps.

Another pros expert argued the maps were only estimates and were inaccurate for cell site anslysis because the maps represented coverage at roof level and no good for ground level (2M).

I have used both maps sucessfully for the defence and prosecution. The essential facts are to say what they can do to inform the court and what they cannot do.

I note the suggestion of the use of circles (ellipsoids?) to define the location of the defendant's mobile, but there is no suggestion as to what data is being used to do that and the elements composite in the data that distinguishes the location data or signalling messages having been sent from the defendant's mobile to the network at the material time.

If the defendant says I was on the street at this corner at the material time using 'the' mobile (admission) then the maps could only add a gloss to either supporting or disproving that claim. But other evidence would be needed along wiht the maps, in addition to radio test measurements.

Greg Smith said...

I advocate the use of best server plots/density maps and single cell predication plots/density maps. I have used both maps successfully for the defence and prosecution. The essential facts are to say what they can do to inform the court and what they cannot do.

I note the suggestion of the use of circles (concentric/ ellipsoids?) to define the location of the defendant's mobile, but there is no suggestion as to what data is being used to do that and the elements composite in the data that distinguishes the location positioning, movement and/or signalling messages having been sent from the defendant's mobile to the network at the material time.

Susan, if you want more details send an email to me. Regards (

Greg Smith said...

Susan, sorry for the double post. I wrote the first post and pushed preview (or thought I had) to check typos and content because the typing screen area is very small. Your blog informed me my post was too large. So I shortened (as you can see) and pressed publish. I am sending this message as I do not know your email address.