How GPS Data Documents Route Deviations in Rideshare Sexual Assault Cases

Published on April 8, 2026

Contact Mark CV Download

What a route deviation actually means in the data

A route deviation is not simply a wrong turn. In engineering terms, it is a measurable departure from the expected path between a pickup and a dropoff, identified through timestamped coordinates that show where the vehicle was and when.

The deviation may be a segment of travel that does not correspond to any reasonable navigation decision, an unexpected stop with duration and location that have no obvious explanation, or a trip timeline that runs significantly longer than the route warrants.

What makes this technically demanding is that not every anomaly in the data is a real deviation. GPS coordinates can look erratic for reasons that have nothing to do with driver behavior.

When a satellite signal arrives at an angle and reflects off a building, a bridge, or a large vehicle before reaching the phone, the resulting coordinate can appear far off the actual route.

A vehicle traveling at a consistent speed does not suddenly appear a quarter mile off route and return within a single data interval. When that pattern appears, it is an error in the signal, not a deviation in the route.

How GPS data gets recorded on a mobile phone

The GPS receiver in a phone does one thing: it receives signals from satellites. It does not transmit location to the satellite.

It calculates position from the signals it receives, and the apps running on the phone ask that receiver for coordinates at intervals, then log them with a timestamp.

Under good conditions, consumer-grade GPS is accurate to roughly ten to fifteen feet (sometimes much better, other times much worse). That accuracy degrades in urban environments with tall buildings, near steel bridges, in tunnels, and around large vehicles.

The degradation happens because signals that should arrive directly from a satellite above instead bounce off nearby structures before reaching the receiver. The coordinate the phone records reflects the reflected path, not the actual position, and that discrepancy shows up in the data.

This is not a flaw unique to any one phone or platform. Every consumer GPS receiver operates under the same physics.

The public has had access to the full-accuracy GPS signal since the late 1990s, when the intentional accuracy restriction that limited civilian GPS to around 150 feet was lifted.

What you get today from a standard smartphone is the same unrestricted signal, interpreted by the same class of receiver, subject to the same environmental interference.

How I identify a deviation in the data

The raw data arrives as a series of coordinates with timestamps. Speed is sometimes included; when it is not, I compute it from the position data.

I can also derive bearing, which tells me the direction the vehicle was heading at each recorded moment. Those three elements together, location, time, and direction of travel, give me enough to reconstruct a route.

The data does not always arrive in a clean format. I have received records as PDFs of spreadsheets that had to be converted before analysis, and complex exports where the information I need is buried within compound data fields rather than laid out in simple columns.

I have built tools over the years to handle these formats because the data you need is not always packaged the way you would want it.

Once the data is structured, I plot it using R, a statistical programming language well suited to this kind of spatial analysis. I can review months of data or just specific dates, whatever you need for your case.

A deviation appears as a departure from the expected path that cannot be explained by normal traffic, navigation, or signal error. When the pattern fits a route a vehicle could not have taken given its speed and the surrounding coordinates, it is an artifact.

When it fits a possible path that differs from what the trip should have been, it warrants a closer look.

The analysis and all supporting plots are written into a formal report using LaTeX, which produces the kind of structured, precise document that holds up as part of a court filing.

GPS uncertainty and what it means for a deviation claim

When I present location data to a jury, I present it as a range, not a fixed point, because that is what it is. The phone recorded where it reported it was, within a margin that depends on the conditions at that moment.

That uncertainty does not erase patterns. A single anomalous point might be dismissed as a signal artifact. A consistent series of coordinates showing the vehicle on a path that does not connect the pickup to the dropoff is not explained by a fifteen-foot accuracy margin.

The question I answer is whether the pattern in the data holds up across multiple points, over time, and in a way that is consistent with actual movement rather than signal noise.

When analyzing the opposing expert’s report, I run the same data and see if my results were the same or different. The appropriate response is to show the full dataset, explain the conditions that affect accuracy, and let the pattern speak for itself. Uncertainty is part of the analysis, not a reason to avoid it.

What a stop looks like in the data

A stop appears as coordinates that cluster in one location over time with no meaningful change in position. The vehicle is not moving, and the data reflects that as a series of points that sit on top of each other or within a very small area for an extended period.

Context is what separates a relevant stop from a routine one. A cluster of stationary points at a busy intersection for thirty seconds looks different from a cluster in an area with no intersection, no destination, and no traffic explanation, held for several minutes.

The location of the stop matters as much as its duration. I plot both.

How this gets presented to a jury

The first thing I do is establish what the data can and cannot show before presenting any findings.

For most cases I start with static maps. A map showing the expected route alongside the actual recorded path gives jurors a visual reference they can interpret without any technical background. I annotate the relevant segments, flag the stops, and walk through the timeline so the sequence of events is clear.

When a static map is not enough to convey what happened, I build an animation. An animation shows the route developing over time with a trail behind it, so jurors can follow the vehicle’s movement as it occurred rather than trying to reconstruct a sequence from a finished map. The timing in the animation is not always to scale with real time, but the order of events and the path taken are accurate to the data.

The goal throughout is to present what the data shows in a way that a jury with no GPS background can evaluate. I am not telling them what happened. I am showing them what the coordinates recorded, what the timeline contains, and where the data is reliable or not reliable.

How to determine if the data is reliable

Be prepared for the data to arrive in a format that requires processing before it is usable. That is normal. What matters is that the underlying records are preserved and produced, not that they arrive in a convenient format.

Whether the data you receive is reliable or not is not something you can determine from the records alone. The same dataset can look reliable or unreliable depending on conditions only an electrical engineer can evaluate. That is the assessment I provide: not just what the data shows, but whether what it shows is worth relying on.

Contact Mark CV Download