The 60Hz Difference: Why Sampling Rate Is the Most Important Number in Vehicle Telemetry

Ask ten telemetry vendors what they measure and you'll get ten identical lists: RPM, coolant temperature, battery voltage, throttle position, oxygen sensor output, fuel trim, wheel speed. The parameters have been standardized for thirty years.

Ask the same ten vendors what rate they sample those parameters at, and you'll get a much more interesting answer. Because that one number — Hz — determines whether the data is diagnostic gold or expensive wallpaper.

What sampling rate actually means

Sampling rate is how many times per second a sensor reports its value. A 1 Hz sensor produces one reading per second. A 60 Hz sensor produces sixty.

That sounds trivial until you remember what's happening inside a running engine. A four-cylinder engine at 3,000 RPM fires 100 combustion events per second. A 1 Hz sampler sees one value per second from anything touching that system — coolant temp, crank position, MAF airflow, injector pulse. It misses 99 out of every 100 events.

That's not a subtle limitation. That's blindness by design.

The Nyquist-Shannon theorem — the foundational result of signal processing — says you need to sample at at least twice the highest frequency you care about, or the signal is unrecoverable. Below that, what you get is aliasing: the sensor reports a smooth, plausible-looking trace that has almost no relationship to the actual waveform.

If the interesting failure signatures live in the 5–30 Hz band — and most mechanical precursors do — you need to be sampling at 60+ Hz to see them. Below that, you're not sampling; you're fabricating.

A practical example: the alternator diode

One of the most common hidden failures in any 10+ year-old vehicle is a partial diode failure in the alternator's rectifier bridge. Six diodes convert the alternator's AC output into the DC your car needs. When one starts degrading, the output voltage develops a periodic ripple — typically at 300–450 Hz, varying with engine speed.

This is diagnostically important because alternator ripple is the earliest detectable sign of:

• Battery-sulfation acceleration (the ripple charges-discharges the battery thousands of times per minute)
• ECU memory corruption risk (modern controllers are sensitive to supply ripple above 400 mV)
• Parasitic drain that your multimeter won't catch

At 1 Hz sampling, ripple is invisible. The voltage reads "14.1 V" once per second. No anomaly, no code, no warning.

At 60 Hz sampling, you see voltage variance. A normal alternator produces readings clustered within ±20 mV. A failing diode produces variance of ±180 mV or more. That's a 9× jump in standard deviation — trivially detectable by the anomaly model.

At 200 Hz sampling, you can reconstruct the ripple waveform itself and identify which specific diode is failing based on the harmonic pattern.

Same alternator. Same sensor. Completely different diagnostic reality.

Why everyone defaults to 1 Hz

Three reasons, none of them good:

Bandwidth costs. At 60 Hz across 200 parameters, a single vehicle produces roughly 10 MB/hour of telemetry data. Across a 10,000-vehicle fleet that's 100 GB/hour, or 2.4 TB/day. Most aftermarket telematics vendors were built when cellular data was $10/MB, so they optimized for the smallest payload possible and the 1 Hz default stuck.

Compute costs. Raw 60 Hz streams have to be processed, not just stored. Anomaly detection on a 60 Hz vibration stream is a fundamentally different engineering problem from threshold-checking a 1 Hz temperature reading. You need actual signal processing — FFTs, wavelet decomposition, recurrent models.

Incentive misalignment. If you're an OEM selling a 3-year warranty, you don't want 48-hour advance warning of bearing wear on a 34-month-old car. It's cheaper to pay for the replacement under warranty than to replace the bearing preemptively and then get asked why the part failed to begin with. 1 Hz telemetry is, from a warranty-economics standpoint, a feature rather than a bug.

Third-party predictive platforms don't have those incentives. Ours certainly doesn't.

What we sample, and why

The Essential tier exists for the same reason most cars ship with OBD-II: the regulatory and basic-visibility floor. The Pro tier is where the category shifts from "observability theatre" to "actually useful." The Enterprise tier is where the work looks more like flight-test instrumentation than automotive diagnostics.

What this means for you

If you're evaluating any telemetry product — ours or anyone else's — the first question to ask is not what do you measure? Ask at what rate?

A 1 Hz product sampling 500 parameters is diagnostically weaker than a 60 Hz product sampling 50. Quality of signal beats quantity of signal, every time. The Nyquist limit doesn't negotiate.

60 Hz is the number that matters. It's the rate at which vehicle telemetry stops being a dashboard ornament and starts being an engineering discipline.