DC Superior Court Judge Rainey Brandt allowed testimony from a cellular analysis expert during a murder trial despite defense concerns of reliability on Dec. 9.
George Sutton, 46, is charged with first-degree premeditated murder while armed, possession of a firearm during a crime of violence, and unlawful possession of a firearm with a prior conviction greater than a year. The charges stem from his alleged involvement in the fatal shooting of 34-year-old John Coleman on May 1, 2023 on the 2000 block of M Street, NE. Coleman suffered from three gunshot wounds to his neck, knee, and thigh.
Steven Kiersh, Sutton’s attorney, challenged the reliability of an expert witness the prosecution planned to call to testify during the trial. Judge Brandt held a Daubert hearing, without the jury present, to determine the admissibility of the expert’s testimony.
At the Daubert hearing, the cellular expert testified–an FBI special agent with the Cellular Analysis Survey Team (CAST). The agent said he used timing advance evidence data, a type of digital forensics evidence provided by cellular networks.
According to the agent, timing advance was not created for law enforcement to use, but for phone carriers to improve users’ experience. The agent explained that cell towers emit radio frequency to estimate how far a phone is from the tower.
The agent said he received T-Mobile records from the lead detective in the case, reviewed the data, and used software to create a map. Then, the agent ensured the records aligned with the tower list generated by T-Mobile. Finally, his report was peer reviewed to ensure accuracy.
Judge Brandt asked how the agent determined the tower list was reliable and the agent said the list updated when changes occurred and he contacted T-Mobile if he had concerns.
According to the agent, the data was reliable because cell networks relied on the information. Additionally, he said the FBI recovered people based on their last known connection to a cell tower. The agent said the FBI first used timing advance in 2009 to locate a missing child in Maryland and since then, law enforcement and the military frequently used the methodology.
The method was limited, said the agent, because the data generated an arc with a device’s possible location but could not determine a specific location on the arc. However, overlapping arcs created more precision.
The agent testified during cross-examination by Kiersh that he was involved in the case since early August 2023, but did not know the number of towers in the area he reviewed. The agent told Kiersh he did not inquire if any changes in tower coverage occurred in the three months between the incident and his involvement.
Kiersh asked if a drive test was conducted during the investigation, when someone is on foot with a radio scanner measuring the strength of the closest tower. The agent said there was no drive test in the case because the timing advance method was superior and provided narrower results.
The agent told Kiersh there was no rate of error for timing advance because the measurement occurred with network activity. Kiersh asked if the method was perfect and the agent said not perfect but one of the best tools. The agent acknowledged a lack of human knowledge or understanding of the data and method could result in errors.
After the agent testified, the prosecution argued they met the Daubert standard to qualify the agent’s expert as reliable. They said the agent received the records from the detective, reviewed them for abnormalities, and there was a peer review of his report.
Kiersh disputed the reliability of the agent and said “I found his testimony troubling.” Kiersh argued the agent unilaterally decided to not conduct a drive test or determine if there were any changes in the cell towers. Kiersh asserted a drive test was the “gold standard” of CAST and the report did not indicate timing advance superseded the method.
Additionally, Kiersh claimed there was a lack of specificity in the method and no rate of error which he asserted an expert was required to establish. The prosecution responded that the rate of error was encompassed in the science itself with the arc of possible locations.
Judge Brandt found the prosecutors satisfied Daubert and allowed the agent to testify as a cellular analysis expert before the jury. The judge said both parties’ arguments ignored a key aspect of the agent’s testimony, that timing advance must be reliable or the networks would fail.
Kiersh then argued to exclude the agent’s testimony because it would be more prejudicial than probative. Judge Brandt denied Kiersh’s request because there was nothing substantially prejudicial and it added to the identity of who prosecutor’s claim drove the suspect vehicle.
After explaining his qualifications to the jury, the agent testified about his report on cell tower data collected in this case.
The prosecutors highlighted two addresses on various maps of cell tower connections. The first location was the address of the shooting, and the second location was labeled as “other.” The agent testified that was the only classification he was given and he was not aware of any relevance to the case.
The first map showed tower connections to a phone prosecutors previously tied to Sutton before noon on the day of the shooting. Two towers near the incident address were highlighted with arcs of possible locations of the phone.
A zoomed in view was displayed of the tower closest to the incident with overlapping arcs less than a block away from the incident.
The prosecution also showed a call record chart of Sutton’s phone and mentioned numbers he contacted around the time of the incident, but did not specify who they belonged to.
The agent noted there was a gap in cell tower data between 11:50 a. m. and 12 p. m. He said that while phones were always searching and connecting to towers, the providers do not always store everything.
Another page from the agent’s report was displayed with several new towers further away from the incident but near the “other” address, showing movement of the device.
A final map was displayed, with more than ten towers highlighted. The agent added a few arcs on the map to show the phone’s movement. The agent concluded the only way the phone could connect to all the towers in the time frame presented would be with a motor vehicle.
During cross-examination, Kiersh asked about the limitations of this type of data. The agent clarified that it was not an exact science, even with timing advance increasing accuracy. The arcs on the maps are a range of where the phone could be and the agent’s report noted “possible location” in his explanation.
The agent further clarified that phones don’t always connect to the closest tower if a further one offers a stronger connection. He concluded that, while cell tower data can’t pinpoint where a phone and its owner are, it can eliminate places it can’t be and suggest possible locations for it.
The Metropolitan Police Department (MPD) lead detective for the case also resumed his testimony with cross-examination from Kiersh. Although the detective did not collect the shell casings from the scene, he confirmed the .45 and 9mm caliber casings could not originate from the same firearm without modification.
The detective confirmed that throughout his interview at the police station, Sutton maintained he did not murder Coleman.
The prosecution also called a litigation technology specialist with the United States Attorney Office (USAO) to testify about a video compilation he created for the case. The prosecutors played the eight minute long video of a car icon moving through a map with surveillance cameras showing the car at various locations.
The trial is scheduled to resume on Dec. 10.
