Last week I posted a poll on LinkedIn with the question:
How can call detail records reveal patterns of driver fatigue or distraction over time?
Here are the potential Poll answers):
- Call/SMS by time-of-day
- Calls overlapping drive time
- Tower handoffs show motion
- Data sessions near incidents
Now let’s take a look at each one of these potential answers.
Call Detail Record (CDR) analysis is one of the most practical ways to evaluate whether a driver may have been fatigued or distracted over time. CDRs are not “phone dumps.” They are carrier-generated business records that typically capture metadata about communications and network events: the time a call or text occurred, direction (incoming/outgoing), duration, the other number involved, and, depending on the carrier and what is produced, cell site/tower and sector information and sometimes data session or other network activity logs.
That limitation is also what makes CDRs defensible when handled carefully. Because CDRs are created and stored by carriers in the ordinary course of business, they can be easier to authenticate than many forms of device content. But because they are metadata, they rarely prove what was on the screen or exactly what the driver was doing with the handset. The value of CDR analysis is in pattern recognition and timeline correlation: showing that communications or network activity repeatedly occurred during likely driving windows, or that a driver’s sleep opportunity was repeatedly constrained based on recurring late-night activity (when corroborated).
What follows expands the four core ways CDRs can reveal patterns, along with practical guidance for attorneys on requesting records, structuring analysis, and presenting results without overstating what CDRs can prove.
First, a quick CDR primer: what you usually get (and what you usually don’t)
While the fields vary by carrier and production format, CDR-related materials commonly include:
- Voice call records: date/time, originating number, terminating number, call direction, duration, sometimes call type/roaming indicators.
- SMS/MMS metadata: timestamps and direction. Content is often not included, and in many situations is not available.
- Cell site/tower + sector information: sometimes listed per event (e.g., tower/sector at call start and end for voice call; tower/sector for SMS). This can support approximate location and movement inference.
- Data session / network event logs: sometimes provided, sometimes not. These can show that the device exchanged data with the network at particular times, but typically do not identify the app or on-screen activity.
- Subscriber/account info: name/address, billing details, sometimes IMEI/IMSI or device identifiers (helpful for attribution and multi-device issues).
What you usually do not get from CDRs alone:
- whether the phone was unlocked or in-hand
- what app was open
- what text content was sent/received (often unavailable)
- whether a call was hands-free vs handheld
- precise GPS location
That’s not a weakness so long as you’re explicit about it. The most credible CDR reports draw bright lines between (1) what the records show and (2) what you infer when you integrate other evidence.
1) Call/SMS by time-of-day: patterns that repeat
Poll answer: Call/SMS by time-of-day
CDR concept: Aggregate call and SMS timestamps over a long period and analyze frequency and duration by hour-of-day, day-of-week, and “known driving windows” (commute, scheduled route, shift transitions).
Why it can reveal distraction over time (under the right circumstances):
A single phone event near a crash can be argued away as unusual. A recurring pattern is harder to explain. If a driver’s CDRs show routine outgoing calls or texting during the same windows when the driver is typically on the road—e.g., weekday mornings 6:30–7:30 a.m. and afternoons 4:30–5:30 p.m.—that supports a habit inference: the driver routinely communicates during travel.
Importantly, the inference isn’t “they were always driving.” The inference is: “their communications frequently occur in time windows that are consistent with commuting/travel.” The strength of this conclusion depends on corroboration (more on that below).
Why it can reveal fatigue risk over time (under the right circumstances):
Fatigue cases often turn on sleep opportunity and circadian timing. If CDRs show repeated late-night activity followed by early-morning activity, especially on work nights, those patterns can be consistent with shortened or fragmented sleep opportunity. CDRs don’t prove sleep deprivation; they can, however, provide objective timestamps that help evaluate whether a person plausibly had time for adequate rest.
For example, a recurring pattern of outgoing calls around 11:30 p.m.–12:30 a.m. followed by outgoing calls around 4:30–5:30 a.m. may be consistent with a compressed sleep window. Standing alone it proves nothing; paired with shift start times, dispatch records, or ELD logs, it can become persuasive.
What corroboration makes it strong:
- Work schedules, dispatch logs, ELD (for commercial drivers)
- Known commute times, jobsite check-in/out, gate logs
- Crash times and roadway conditions (nighttime, monotony, etc.)
- Testimony (driver, employer, co-workers) about typical routines
Pitfalls and how to avoid them:
- Shared plans / shared devices: Confirm who controlled the phone when. Subscriber ≠ user.
- Sparse records: Some drivers rarely call/text; absence of evidence is not evidence of absence.
- Weekend vs weekday differences: Keep segments separate; patterns can differ dramatically.
2) Calls overlapping drive time: the most intuitive timeline overlay
Poll answer: Calls overlapping drive time
CDR concept: Identify calls whose start and end times overlap with a period you can independently establish as driving, or at least vehicle movement. This is often the cleanest form of CDR-based distracted-driving analysis because voice calls are timestamped and have durations.
Distraction inference (when correct):
If a driver is on a call during a time when they’re verifiably driving—especially if it’s a pattern over weeks/months—CDRs can support:
- the driver’s attention may have been divided (cognitive distraction), and/or
- the driver may have been handling the phone (depending on other evidence, not the CDR alone).
Even if the jurisdiction distinguishes handheld vs hands-free, calls can still matter. CDRs can help show the activity existed during driving; the legal characterization, hands-free permitted or not, depends on local law and the rest of the evidence.
Fatigue inference (secondary but sometimes relevant):
Calls can also be used to map rest opportunities:
- Long calls during expected sleep windows can indicate time awake.
- Repeated patterns of late call and early departure can support a chronic fatigue theory—particularly in commercial contexts with long duty cycles.
How to make the driving window defensible:
CDR analysis lives or dies on whether you can credibly say “more likely than not, the driver was driving at that time.” The best sources to establish this are independent records, such as:
- vehicle telematics (speed, ignition, trip start/stop)
- ELD duty status changes and driving segments
- dispatch pickup/delivery timestamps
- toll tags, bridge crossings, ALPR hits
- receipts (fuel, meals) with timestamps
- security camera timestamps
- witness statements (we saw them leave at X)
Once the window is established, you can overlay the call. That overlay can be presented with simple exhibits: a timeline with a “driving bar” and a “call bar,” showing overlap.
Pitfalls and how to avoid them:
- Incoming calls: An incoming call record doesn’t mean the driver answered. You need duration/connection indicators; some records show zero-duration attempts.
- Voicemail and routing: Distinguish connected calls from forwarded/voicemail events if the carrier format shows it.
- Time zone normalization: Make sure all timestamps are normalized (carrier logs may be UTC or local; documents can mix formats).
3) Tower handoffs show motion: using cell site/sector carefully
Poll answer: Tower handoffs show motion
CDR concept: Some CDRs include cell tower and sector at call start/end, during data sessions, or per SMS event. When those tower/sector fields change in a way consistent with geographic travel—especially along a known route—you can sometimes infer that the device was moving.
This can be a valuable bridge between “a call occurred” and “the phone was likely traveling.” But it must be handled with care.
How it supports distraction patterns (when correct):
- If the driver is on a call and tower/sector fields change across the call duration in a way consistent with moving along a route, you can argue the device was not stationary.
- Over time, repeated calls with tower progression during commuting windows can support a pattern of calls while traveling.
How it supports fatigue patterns (when correct):
- Tower sequences can help validate that a person was traveling at unusual hours repeatedly (e.g., consistent early-morning highway travel), which can support fatigue-risk arguments when paired with work/rest constraints.
Core caution: cell site ≠ precise location
Cell site/sector information is approximate. Phones don’t always connect to the closest tower. Network load, terrain, building density, and carrier optimization can cause “unexpected” tower selection. Therefore:
- Treat tower data as consistency evidence, not pin-point proof.
- Avoid claiming exact speed, lane, or exact position from towers alone.
- Use tower analysis to test whether a narrative is plausible (the phone was generally progressing south along I‑10.), not to claim millimeter-level precision.
What corroboration makes tower analysis stronger:
- Known route or route options (from dispatch, GPS, receipts, or testimony)
- A sequence of events relatively close in time (SMS events within minutes, or call start/end towers with a long duration call)
- Maps
- Vehicle data (speed/heading) if you have it
Common pitfalls:
- Over-interpreting a single tower: One tower hit is weak; sequences are stronger.
- Urban density: Many towers in a small area can make movement appear more dramatic, or not meaningful.
- Rural coverage: Fewer towers mean bigger coverage areas; movement inference may be coarse.
4) Data sessions near incidents: timing evidence, not app proof
Poll answer: Data sessions near incidents
CDR concept: Some productions include data session records or other network activity logs. These can show that the device exchanged data with the network at or near certain times.
Why it can indicate distraction (under the right circumstances):
If you see one or more data sessions in the minutes leading up to a crash or near-miss, that can be consistent with:
- message sending/receiving,
- social media refresh,
- navigation activity,
- streaming adjustments,
- or any number of background processes.
Because of that ambiguity, the defensible claim is typically about timing correlation: “the device was active on the network very near the incident,” not “the driver was scrolling an app.”
Over time, however, repeated data sessions during established driving windows can support a broader pattern claim: the device frequently interacts with the network during trips. That can matter where the legal or safety argument hinges on divided attention generally, or where it corroborates other indicators.
Ultimately, the data usage shown on the CDRs can be used to strengthen the argument to demand the associated device be forensically examined. The only way to concretely show whether the data usage was active use by the user or passive use by the background processes is to examine the device itself.
Why it can indicate fatigue risk (secondary):
Regular late-night data activity can serve as another objective indicator that a person may have been awake at times they might otherwise claim to have been sleeping. Again: supportive, not dispositive.
How to present data-session evidence responsibly:
- Use careful phrasing: “network activity consistent with data usage occurred at X time.”
- Discuss alternative causes: background updates, push notifications, system services.
- Strengthen with context: was the phone stationary? was there a call simultaneously? do other records show driving?
Pitfalls:
- Treating data sessions as “screen touches” (active use by the user).
- Ignoring that some carriers batch or log sessions in ways that don’t correspond neatly to human interactions.
Tips for Attorneys
1) Request the right fields—explicitly
Carrier returns vary. If you want tower/sector, ask for it. If you want data session logs, ask for them. Consider requesting:
- voice CDRs with start time, end time, duration, direction
- SMS/MMS metadata (time, direction, other number)
- cell site/tower + sector for each event (and for call start/end if available)
- data session/network activity logs (if available)
- subscriber/device identifiers (IMEI/IMSI where available)
- preservation letters early (where appropriate)
- Email me at ben@braveinvestigations.com for sample language.
2) Don’t over-claim what CDRs prove
Avoid statements like “the driver was texting” unless you have corroboration beyond CDR metadata. More defensible:
- “SMS events occurred at these times.”
- “Those SMS events coincide with an established driving interval.”
3) Use repeatability to your advantage
Jurors and adjusters understand habits. A single overlap is arguable. Ten overlaps during commute windows is a pattern. Visuals help:
- a calendar heatmap of call/SMS counts by hour
- a timeline showing driving bars and event ticks across multiple days
4) Anticipate attribution challenges
Be prepared to address:
- family/shared phones
- employer-issued phones used by multiple employees
- call forwarding or voicemail artifacts
- the possibility of a passenger using the device
Sometimes the best approach is to treat CDRs as one piece in a mosaic with testimony and other business records.
5) Prepare clean demonstratives
The easiest persuasive exhibit is often:
- Driving window: 7:12–8:03
- Call: 7:41–7:52
- Overlap: 11 minutes
Then repeat for multiple dates to show pattern.
6) Bring in an expert early
CDR productions vary by carrier and can be easy to misread.
· time zones
· connected vs. attempted calls vs. forwarded to voicemail
· tower/sector fields
· what “data sessions” do and don’t mean
Retaining a qualified CDR/cell site analyst early, ideally before subpoenas go out, can help you request the right records the first time, build a defensible driving timeline for overlays, and avoid overstatements that can become problems in deposition or at trial.
Conclusion
Call Detail Record analysis can reveal patterns of driver
fatigue risk and distraction over time by documenting when
communications and network events occurred and, in some cases, providing coarse
indicators of movement through cell site/sector fields. The most
defensible findings come from (1) aggregating CDR events to identify repeatable
timing patterns and (2) overlaying those events onto independently established
driving and rest windows. CDRs rarely answer every question on their own, but
they often provide the objective timestamp backbone that lets the rest of the
case evidence snap into a coherent story.
If you have any questions please don’t hesitate to contact
me at ben@braveinvestigations.com
or through the contact form on this page.