Flat list for a graph

Data with parent-child relationships is shown as rows, and users rebuild the structure in their heads.

A table with expandable rows can work for shallow structures and precise quantity comparisons. A graph helps with branching and traversal. A regulatory response deadline concerns available records and retrieval; it does not prescribe a graph interface.

• Shown as rows, the same identifiers repeat across many rows.
• Users draw diagrams on paper to see the structure.
• Simple trace questions take long investigations.
• Under the FDA’s Food Traceability Rule, covered firms must provide records to the FDA within 24 hours of a request or within a reasonable time the FDA agrees to.
• Bills of materials, approval chains and dependency maps contain relationships, but their structure and the user’s questions differ. Choose a graph, tree or table accordingly.

• The same identifiers repeat across many rows
• Users draw diagrams on paper
• Simple trace questions take long investigations
• A flat trace report grows too long to read across merges, splits and many transformation steps

Tables are the default component. Lists and tables come with every component library. A graph view has to be designed and built.

Data stored as rows. Relationships are stored as rows in a database, and the screen shows them the way they are stored.

• Slow answers. Simple trace questions take long investigations.
• Errors in the investigation. People rebuilding a structure in their heads miss branches, which matters most in a recall.
• Regulatory risk. Where records must be produced within a fixed time, an investigation that takes days is a compliance problem.

Where it shows up

• Lot genealogy map
• Traceability design
• The report is too long to be read
• Traceability falls short of a regulated industry’s requirements
• Legacy application modernization

Sources: FDA, FSMA Final Rule on Requirements for Additional Traceability Records for Certain Foods