Sensor and time-series data quality

Find and fix bad sensor data in your browser

Clean signals, trusted decisions.

See the gaps, spikes, flatlines and drift in a time series. Preview each repair. Then write the rule that catches the problem next time.

No upload. No account. Free during early access.

A teal signal line on a dark grid breaks and is mended by small amber stitches.
  • Gap
  • Spike
  • Flatline
  • Drift
Sample data. Issues marked by Signal Check.

Four ways sensor data goes wrong

Most bad readings fall into four patterns. Sigmend checks for all four.

Gaps

Readings stop arriving. A logger drops out, a network link fails, or a historian skips a period. Your chart draws a straight line over missing time.

How gaps are found →

Spikes

One reading jumps far from its neighbours and comes straight back. It can set off an alarm the process never earned.

How spikes are found →

Flatlines

The same value, again and again. The sensor may be stuck, or the value may be held after a fault. The process looks calm when it may not be.

How flatlines are found →

Drift

The readings slowly move away from where they should be. Often the cause is calibration. Nothing looks wrong on any single day.

How drift is found →

Check, repair, prevent

  1. Check.

    Load a CSV export or the sample. Signal Check marks every gap, spike, flatline and drift section on a chart. Each one comes with a plain-language explanation.

  2. Preview the repair.

    For gaps and spikes, Repair Preview shows a before and after view. You accept or reject each change. Then you download the repaired CSV and a log of every change.

  3. Write the rule.

    Rule Builder turns a limit into a readable rule plus JSON and YAML you can paste into your pipeline. You can test the rule on the data you have loaded first.

Read the methods and defaults →

A teal signal line loops entirely inside a closed boundary sealed by an amber stitch.

Privacy

Your file stays in your browser

No upload. No account. Your browser reads the file you pick, and the file never leaves your device. The checks are plain JavaScript with fixed, documented rules. They make no network calls.

Built for people who depend on sensor data

  • Operations and reliability engineers. Catch a stuck or drifting tag before it hides a real fault or delays maintenance.
  • Data engineers. Keep bad readings out of pipelines and dashboards, using rules you can export as JSON or YAML.
  • Analysts. Show what changed in a dataset and why, with a before and after view and a change log.
  • IoT product teams. Run a fast check on field-test telemetry for dropouts and spikes.

For teams at manufacturers, energy and utility operators, building-management firms and IoT product companies.

Plans

Early access: free now

Available now

You get all three tools: Signal Check, Repair Preview and Rule Builder. No account is needed.

Try with sample data

Paid plans: coming

Coming

They are not available yet. There are no prices to show.

See plans and pricing →

What we plan next

  • More file formats, starting with JSON Lines.Planned
  • Checks across several sensors at once.Planned
  • Saved rule sets.Planned
  • Connectors to common data sources.Planned

Planned means we intend to build it. It does not mean there is a date. Full roadmap →

Join early access

Get a short email when new checks, formats or plans arrive.

We use your email only for Sigmend early-access updates. See our Privacy page.