Main menu

From Manual Checks to Machine Confidence

How MassDOT Modernized Data Validation...

State Departments of Transportation often spend countless person-hours manually validating GIS datasets. Unfortunately, it’s an approach that doesn’t scale as data volumes grow. In Massachusetts, for example, MassDOT wanted to fuse third-party INRIX traffic segments (TMCs) with its official road inventory for planning and reporting.

Under the old system, this “conflation” process (or “data fusion,” as we call it) was prohibitively slow: aligning INRIX and LRS data took four full-time staff about a year to cover only 70% of the network. Such labor-intensive checks made it nearly impossible to update data quickly or reliably.

Beneath that year-long effort lay the real issue: reconciling two fundamentally different data models.

The Pain of Manual Data Fusion

Before automation, MassDOT’s data analysts wrestled with many technical mismatches when merging the TMC data into the Linear Reference System. For example:

  • Different network coverage: Some INRIX TMC segments lie outside the areas MassDOT’s road network covers (and vice versa). Extra logic was needed to decide when a TMC belonged in the LRS or when parts of the LRS had no matching TMC.
  • Misaligned linework: Like with almost all linear network inventories, the INRIX centerlines didn’t match MassDOT’s geometry exactly because each dataset was built for different purposes, scales, and usage patterns. In practice, this meant analysts had to manually apply “fuzzy” buffer tolerances to align road segments before any real conflation could happen.
  • Conflicting segmentation: INRIX breaks roads at every intersection, whereas MassDOT’s LRS is an end-to-end route depiction so it’s segmented infrequently. That forces many-to-one matching between multiple short TMC pieces and single longer LRS routes.
  • Variant representations: Ramps, divided highways and junctions often appear in INRIX differently than in MassDOT’s data, requiring case-by-case adjustments.
  • Inconsistent naming: The two datasets used different naming conventions, so even identical road names might not match without manual mapping or “fuzzy” attribute rules.

Each of these differences made purely manual QA inefficient and error-prone. 

As we alluded to earlier, MassDOT’s attempts without specialized tools took 12 months and four people to achieve only ~70% data fusion coverage. With updates to INRIX due twice a year, it was clear a new approach was needed.

Enter Rule-Based Automation with 1Integrate

MassDOT’s solution was to apply 1Spatial’s 1Integrate rules engine (along with 1Data Gateway) to automate the data fusion.

In practice, this meant codifying the expert logic (buffer distances, name-matching rules, segment merging, etc.) into an automated workflow. The system follows a staged process: it first applies validation rules to each incoming dataset, then standardizes every input to a common schema, and finally performs change-detection/matching to align features from the INRIX feed with MassDOT’s LRS.

1Integrate took the place of manual checks at each step. For example, it automatically ran quality checks and standardized projections and attribute formats, eliminating the need to configure custom scripts for each dataset. It then used buffered spatial joins and name-matching rules to detect where INRIX segments aligned with the official network. This rules-based automation captured most of the “easy” matches without human intervention.

As a result of this automated process, MassDOT saw a dramatic speedup. In one test run, 1Integrate “processed the entire state within two hours,” automatically conflating 91% of the TMC segments. MassDOT staff were able to handle nearly all of the network through a repeatable machine workflow, instead of the year-long manual effort before.

Results: Faster, More Accurate, More Confident

The business impact for MassDOT was immediate. A job that used to take a year of tedious checks now runs in a couple of hours, enabling updates on a much tighter schedule.

Frequent INRIX updates (which were once impossible to ingest quickly) can now be processed promptly for planning and reporting. And because the same rules engine runs every time, the process is consistent and traceable, dramatically reducing the risk of human error. Most importantly, this automation built confidence in the data.

With 1Integrate’s clear, rule-based quality checks, MassDOT can trust that mismatches are caught systematically. In fact, the agency notes that the new workflow “enabled MassDOT to confidently integrate” the third-party traffic data into its official LRS. In other words, they finally have a validated, merged dataset they can rely on for analysis, performance monitoring and compliance reporting, rather than worrying that manual mistakes have crept in.

From Effort to Assurance

By swapping manual inspections for an automated, rules-driven data fusion engine, MassDOT turned a year-long slog into a matter of hours. The combination of 1Spatial’s automation platform (1Integrate) and its data-pipeline tools now delivers faster updates, higher accuracy and an auditable data lineage.

For transportation agencies, this shift from “manual checks to machine confidence” means more reliable GIS data, less risk, and more time to focus on engineering and planning rather than spreadsheets.

Want to learn more? Read the MassDOT case study or contact us for a demo of 1Integrate’s transport data solutions. See how your team can boost data quality and trust through automation and rules-based governance.