Crash data retrieval equipment: what it reads, who may use it and how to judge results
How event data downloads work, what the law allows and how to check a report
Safety Behind the Wheel Foundation built this guide from the federal event data recorder rule, the federal Driver Privacy Act, a National Highway Traffic Safety Administration (NHTSA) investigation program and published research on vehicle data and collision reconstruction. Every source is listed at the end.
What crash data retrieval equipment does
Many vehicles carry an event data recorder. It is a function in the vehicle’s electronics that saves a short record of speed, braking and other data around a collision, so someone can read it afterward [6]. The record stays in the vehicle until a person with the right equipment and access retrieves it.
Federal rules require a notice about this in the owner’s manual of a covered vehicle. Reading the data takes special equipment, plus physical access to the vehicle or its recorder. The maker and others, such as law enforcement, can read it if they have both [1].
That equipment is crash data retrieval equipment. A typical setup pairs hardware that connects to the vehicle or recorder with software that turns the saved file into a readable report. A trained analyst runs it. Many are crash reconstructionists, who work out how a collision happened from physical evidence and data.
Typical users include police collision units, other investigators, road agencies, insurers and private reconstruction firms. Their shared goal is safety management: learning what happened and what would prevent a repeat. Equipment comes from authorized professional instrument vendors and technical service providers. A group that rarely needs this work may do better to hire a trained provider.
The federal rules behind the tools
The main federal rule is 49 CFR Part 563 (Title 49 of the Code of Federal Regulations), the EDR rule. It covers passenger cars, SUVs, vans, pickups and buses built on or after September 1, 2012, with a weight rating of 8,500 pounds or less, if they have an EDR [6]. The weight rating is the most the maker says the vehicle may weigh when loaded. The rule does not make an EDR mandatory. It sets standards for those that exist, so the data help collision investigations and show how safety equipment performed [6].
Four parts of the rule matter most for equipment:
- Tool availability. Each maker must make sure, through a licensing deal or other means, that a retrieval tool can be bought commercially. It must be on sale within 90 days after the vehicle’s first sale for purposes other than resale [2]. The duty covers the data the rule requires. Extra data may need other methods [2].
- Data accuracy. Required data must meet set ranges, accuracy and resolution. The rule’s table allows a margin of plus or minus 1 kilometer per hour for indicated speed and plus or minus 10 percent for longitudinal delta V [7]. Indicated speed is the speed the vehicle’s own systems report. Delta V is the change in a vehicle’s speed during impact.
- Crash tests. In tests, required data must still be readable with the maker’s method for at least 10 days afterward, and the record must be marked complete [8].
- Record length. Today, covered recorders save 5 seconds before an event at two readings per second [9].
That is set to change in stages. The amended rule will move required speed, pedal and brake data to 20 seconds at ten readings per second. The change starts with 25 percent of each maker’s covered vehicles built from September 1, 2028, and reaches all vehicles built on or after September 1, 2031 [9]. Small volume makers and makers of altered or multistage vehicles get until 2032 or 2033 [9]. Tools and software will need to read both formats during the changeover.
When a recorder saves data and when it does not
An analyst’s first question is whether the recorder had any reason to save a record. It does not tape a whole drive. It saves data only for an “event,” which is a collision or other occurrence that changes the vehicle’s speed by 8 kilometers per hour (about 5 mph) within 150 milliseconds. A deployed air bag or other one time restraint also counts [6].
A frontal air bag deployment must be saved, and its memory is locked so later events cannot write over it. Other events can fill up to two memory slots. If both are full, the maker may let a new event replace an older non deployment record, or may skip the new one [6].
So the order of events matters. A minor bump after a collision can erase an earlier record that was never locked. This is one reason to leave a vehicle untouched until someone decides whether to retrieve its data.
What the equipment cannot read
Expect these gaps:
- Vehicles outside the rule. Vehicles built before September 1, 2012, and those rated above 8,500 pounds fall outside Part 563 [6]. They may record less, record differently or record nothing. The tool duty sits inside the same rule, so it does not reach them [2, 6].
- Collisions below the trigger. No event means no record [6].
- Overwritten records. An unlocked record can be replaced by a later event [6].
- Personal details and location. The owner’s manual notice says no personal data, such as name, gender, age or accident location, are recorded. It describes a short window of vehicle data, typically 30 seconds or less [1]. The rule describes an EDR as saving vehicle data over time, such as speed and delta V [6].
Who may use the equipment: consent, court orders and other limits
The Driver Privacy Act of 2015, a federal law, says data in an EDR belongs to the vehicle’s owner, or to the lessee if the vehicle is leased. That holds no matter when the vehicle was built [3]. It also bars anyone else from accessing the data unless one of five conditions is met [3].
Comparison compiled by Safety Behind the Wheel Foundation
Five legal routes to read event data recorder data
| Route | What the law says | What to watch |
|---|---|---|
| Court or other authority | A court, or another judicial or administrative body with jurisdiction, approves the retrieval | Ask what the order covers [3] |
| Owner or lessee consent | Consent in writing, electronically or by recorded audio, for any purpose including diagnosis or repair, or a subscription agreement that explains how the data will be read and used | Read the scope before you sign [3] |
| Federal safety inspection | Retrieval during an investigation or inspection authorized under two sections of Title 49 | Personal details and the vehicle identification number (VIN) may not be disclosed with the data. Only the VIN may go to the certifying manufacturer [3] |
| Emergency medical response | Retrieval to decide whether emergency help is needed or to speed it | Tied to that purpose [3] |
| Traffic safety research | Retrieval for research | Personal information may not be disclosed [3] |
The Act’s list names no separate route for police or insurers. Their access generally rests on consent or on authorization from a court or another authority [3]. A trained analyst treats this as step one, before any cable is connected.
For work vehicles and leased cars, check who is the owner or lessee on the title or lease. That party holds the rights [3]. Your state may add rules about consent, access and what insurers may ask. State laws differ and change, so check yours before relying on any summary, including this one.
How a trained analyst runs a retrieval
A careful retrieval follows a routine like this one.
- Confirm authority. Identify the owner or lessee and keep the consent form or order with the file [3].
- Record condition and custody. Photograph the vehicle, note the VIN, and write down who has held it and where it has been stored.
- Match the tool to the vehicle. Check that the tool and its software version support that make, model and model year.
- Connect and read as the tool directs. Log the date, time, tool, software version and operator.
- Note the power up counts. The rule defines one count at the crash and one at the download, so the gap shows how many times the recorder was powered between the two [6].
- Save the original file unchanged. Work from a copy.
- Check the record itself. Look at how many events were saved, which are locked, and whether the record is marked complete [6, 8].
- Test the timeline against other evidence. Compare it with the scene, the damage, any video and the other vehicle.
- Report the limits. State what the data show and what they cannot show.
Reading the report: what the numbers mean
A report looks exact. Treat it as strong evidence with known limits.
Comparison compiled by Safety Behind the Wheel Foundation
What to check in a retrieval report
| What to check | What it tells you | What to watch |
|---|---|---|
| Indicated speed | The vehicle’s own reading. The rule allows a margin of 1 km/h on the recorded value [7] | It is not an independent measurement |
| Reading spacing | Two readings per second now, ten per second in newer vehicles [9] | A fast change between readings can hide |
| Delta V | Change in speed over about the first quarter second of impact [6] | It measures severity, not cause |
| Event status | Whether the record is locked and marked complete [6, 8] | An unlocked record may have replaced an older one |
| Power up counts | Counts at the collision and at the download [6] | A gap means the recorder was powered again before the download |
Recorded values also help test other methods. Federal crash investigators have estimated delta V from vehicle damage with a computer program [5]. One study compared those estimates with recorder values in 121 cases from 2000 to 2003 [4]. The damage based estimates ran 23 percent low on average [4]. The miss varied by vehicle type: 3 percent low for pickups and 31 percent low for front wheel drive cars [4]. A later study of side collisions found the approach ran about 12.9 percent high for cars struck by cars [5]. These gaps show why analysts want measured data, and why no single method gets blind trust. Safety Behind the Wheel Foundation encourages families and employers to read every report the same way.
Some vehicles record more than the minimum. In one study, certain vehicles from model year 2009 on also saved steering wheel angle. Researchers used it to rebuild how vehicles moved before a collision. Their model matched the recorded motion in general, but not when the vehicle left the paved surface or when stability control was active [10]. Treat those situations with extra care.
Common mistakes
- Treating indicated speed as exact travel speed.
- Using a tool or software version that does not support the vehicle.
- Reading the data without clear authority.
- Waiting until the vehicle is repaired, sold or scrapped.
- Treating the report as proof of fault.
Cross checking with the scene and other evidence
Recorded data answers questions about one vehicle. The scene answers others. Technicians in the National Highway Traffic Safety Administration’s national crash investigation sample document skid marks, fluid spills and struck objects, and measure vehicle damage [11].
Video can help too. In one published case study of a fatal collision, a speed worked out from video frames matched the recorder’s speed within an average of about 1.4 mph [12].
When only one vehicle has a record, work can still go on. One research team built a method that uses data from a vehicle with a recorder to estimate speeds for a vehicle without one. It also worked when the recorder data were partly lost [13]. Recorded data show what a vehicle did. They do not show why, and they do not decide fault.
Choosing equipment or a provider
The agency, employer or company that does the investigating chooses the equipment. Residents and parents rarely buy it. School boards, councils and small fleet owners may be asked to fund it or hire a provider. These questions help any buyer.
Upkeep matters. Software needs updates as new models and formats appear, and operators need refresher training. Logging tool and software versions on every job lets a later reviewer repeat the work.
What drivers, families and employers can do
Many people will face a request for a download. These steps help:
- Keep the vehicle as it is. Repair, sale or scrapping can remove the recorder, so ask before any work starts.
- Ask three questions. Who wants the data, why, and on what authority?
- Read any consent form for scope. Does it cover one download or ongoing access [3]?
- Request the full report. Ask for the original file, a printed copy and a note of the data’s limits, then add it to your crash records.
- Ask how to request a review. If a report seems wrong, ask the agency that wrote it how to raise a correction.
- Employers: build it into your process. Decide who is the owner or lessee of each vehicle and who may approve a download [3]. Our work crash and near miss investigation guide shows where this fits.
- Get advice when stakes are high. If someone was badly hurt or fault is disputed, talk with a lawyer before you consent.
Frequently asked questions
Can I buy a retrieval tool for my own vehicle?
Federal rules require makers to see that tools are sold commercially [2]. But buying a tool is the easy part. Reading data well takes training, current software and access to the right module. The data from a vehicle you own are yours, but reading anyone else’s vehicle needs a legal route [3].
Does a download erase or change the record?
Careful analysts take steps to read the file without changing it, and they keep the original. Ask any provider to describe those steps. The record also stores power up counts at the crash and at the download, so a reviewer can see how many times the recorder was powered in between [6].
Can police read my vehicle’s data without my consent?
Federal law allows retrieval with court or other authorization, among other routes [3]. Your state may add rules. If police ask for data, you can ask what authority they rely on. A lawyer can explain your rights.
Will the data show who caused the accident?
No. The required data cover one vehicle: speed, pedal use, belt status and similar items [9]. They do not show the other driver, the signal or the reason behind a choice. Analysts combine them with scene evidence and other records [11].
What if my vehicle is older, or a heavy truck?
Vehicles built before September 1, 2012, and those rated above 8,500 pounds fall outside Part 563 [6]. They may record less or record differently, and the commercial tool duty does not cover them [2, 6]. Ask a trained provider what a specific vehicle can offer.
The bottom line
Recommendation from Safety Behind the Wheel Foundation
Safety Behind the Wheel Foundation urges families and employers to keep a vehicle intact after a serious collision and to ask who wants its data and why. Crash data retrieval equipment turns a short snapshot inside a vehicle into usable evidence. The tool matters less than the process around it: clear authority, the right tool for that vehicle, careful handling and a plain statement of limits.
For families and employers, the best moves come early. Keep the vehicle intact, ask who wants the data and why, and request the full report.
Related articles
Sources and further reading
- 1U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR 563.11, Information in owner’s manual. Up to date as of September 24, 2026. ecfr.gov/…/section-563.11
- 2U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR 563.12, Data retrieval tools. Up to date as of September 21, 2026. ecfr.gov/…/section-563.12
- 3Cornell Law School, Legal Information Institute. 49 U.S. Code 30101, Purpose and policy, including the statutory note for the Driver Privacy Act of 2015 (Pub. L. 114-94, div. B, title XXIV, subtitle C). No update date shown; accessed October 2026. law.cornell.edu/uscode/text/49/30101
- 4Niehoff P, Gabler HC. The accuracy of WinSmash delta-V estimates: the influence of vehicle type, stiffness, and impact mode. Annual Proceedings of the Association for the Advancement of Automotive Medicine, vol. 50, pp. 73-89 (via PubMed). Published 2006. Data year: NASS/CDS 2000-2003. pubmed.ncbi.nlm.nih.gov/16968630
- 5Johnson NS, Gabler HC. Evaluation of NASS-CDS side crash delta-V estimates using event data recorders. Traffic Injury Prevention, vol. 15, no. 8, pp. 827-834 (via PubMed). Published 2014. Data year: not stated in the abstract (NASS-CDS side crashes). pubmed.ncbi.nlm.nih.gov/24433050
- 6U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR Part 563, Event Data Recorders (sections 563.1 to 563.5 and 563.9 opened). Up to date as of October 1, 2026. ecfr.gov/…/part-563
- 7U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR 563.8, Data format (Table III). Up to date as of September 24, 2026. ecfr.gov/…/section-563.8
- 8U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR 563.10, Crash test performance and survivability. Up to date as of September 18, 2026. ecfr.gov/…/section-563.10
- 9U.S. Government Publishing Office and Office of the Federal Register, Electronic Code of Federal Regulations. 49 CFR 563.7, Data elements (89 FR 102832, as amended at 91 FR 28440). Up to date as of October 2, 2026. ecfr.gov/…/section-563.7
- 10Kusano KD, Sherony R, Gabler HC. Methodology for using advanced event data recorders to reconstruct vehicle trajectories for use in safety impact methodologies (SIM). Traffic Injury Prevention, vol. 14, suppl., pp. S77-S86 (via PubMed). Published 2013. pubmed.ncbi.nlm.nih.gov/23905880
- 11National Highway Traffic Safety Administration. Crash Investigation Sampling System. No update date shown; accessed October 2026. nhtsa.gov/…/crash-investigation-sampling-system
- 12Epstein B, Westlake BG. Determination of vehicle speed from recorded video using reverse projection photogrammetry and file metadata. Journal of Forensic Sciences, vol. 64, no. 5, pp. 1523-1529 (via PubMed). Published online March 28, 2019. pubmed.ncbi.nlm.nih.gov/30919949
- 13Han I. Optimization analysis in reconstruction of vehicle collision accidents with deficient EDR data. Forensic Science International, vol. 283, pp. 103-110 (via PubMed). Published December 14, 2017. pubmed.ncbi.nlm.nih.gov/29278821
This article reflects the safety priorities of reliable vehicle data evidence and careful, fair collision investigation. It is general education, not legal, insurance, medical or financial advice. Federal and state rules change, so verify details with official sources. For your own situation, contact the investigating agency, your state insurance department or a qualified crash reconstruction provider, and speak with a lawyer licensed in your state before you consent to a download. Reviewed October 2026.
Cite this article: Safety Behind the Wheel Foundation. “Crash data retrieval equipment: what it reads, who may use it and how to judge results.” drivewithcare.org, https://drivewithcare.org/crash-data-retrieval-equipment-explained/. Reviewed October 2026.
Special thanks to our contributors
We appreciate the professionals, experts, and community members who share their knowledge and perspectives to help make this resource better. Your insights are always welcome and appreciated.
Have professional expertise, firsthand experience, or a perspective that could make this resource better?
We welcome your input. Your knowledge and experience can help us make this resource more useful, accurate, and meaningful to the people who rely on it.
Thank you. Your contribution is on its way.
We’re grateful you took the time to share your knowledge with us. Our team reviews every contribution with care. If it helps improve this resource, we’ll update the page and recognize you the way you chose.
Keep learning
Safe Driving Resource Center
Guides, self-checks and articles for drivers, parents of new drivers, and families recovering after a crash.
