Automated Data Validation

Web browser application improves operational data quality & monitoring site uptime

illustration of laptop and web browser with KISTERS Data Validation portal app. Validation algorithms mark data points as incidents or warnings, when data values fail specific rules, thresholds or conditions.

Human-machine teamwork maximizes resources

As monitoring networks expand and collected data grows exponentially, manual data validation becomes impossibly time-consuming.

Affirm that quality data supports your operations and forecasts.

KISTERS automated data validation portal app directs users to specific time series & monitoring stations that need attention.

Algorithms and machine-learning models screen and cleanse data faster — and with fewer errors.

Improve generation forecasts with complete and correct hydromet measurements.

Key features

Over the loop paradigm

The over the loop paradigm empowers users to intervene in algorithm-driven / machine-based processes. In contrast, the out of the loop paradigm lets artificial intelligence run without human interaction.

Workflow scheduler

Users define and manage their data validation workflow using an intuitive tool embedded in the app. Tasks in the schedule trigger data validation processes. Apply to both high-resolution and aggregated data.

Rule Flexibility

Users can write their own rules, simple or complex.

Instantly deploy rules.

Edit and enhance rules without impairing existing functionality.

Address questions about

Are data values present?

Data Consistency

Quickly identify data gaps. Know when a sensor fails to transmit data. Then triage maintenance and repair needs. Know which sensors or monitoring sites need immediate attention.

Are data values plausible?

Data Validation

Instantly identify if data values are plausible using time series, statistical values or physical constraints from neighboring monitoring stations.

Other references for plausibility include satellite data or
radar-derived precipitation estimates.

Can I fill data gaps or replace implausible values?

Data Estimation

Easily replace missing data or low-quality data with plausible data values from neighboring sensors.

Alternatively, fill with data derived from a variety of interpolation methods or reconstructed from models.

In addition to smoothing noise, remove extreme spikes & outliers.

photo backside of male public works employee wearing a yellow hard hat and orange heavy duty rain jacket to respond to field work during a rain storm

photo of rain falling on a clear umbrella with data overlay