NEO EBR: electronic batch records for regulated manufacturing

An electronic batch record (EBR) is the digital form of the master and executed batch record: the document that instructs how a batch is made and captures what actually happened while it was made. It replaces the paper batch record, and in regulated manufacturing it is the primary evidence a batch was produced according to its approved process.

The distinction that matters is not the format. A scanned PDF is digital and still passive. An EBR is executable: it holds the procedure as a sequence the system enforces, checks entries against their limits at the moment they are made, and requires a second-person verification where the procedure calls for one.

NEO EBR is that record for GxP environments, inside manufacturing execution. It runs procedures and work orders on the floor, reads process values from the historian rather than asking an operator to transcribe them, and produces a record designed to be reviewed by exception rather than line by line.

Digitising "the batch record" is really four documents, and a project that scopes only the first one tends to leave the paper in place. In a typical pharmaceutical operation the set is:

• The manufacturing instruction — the procedure itself, broken into stages such as weighing, compounding, primary and secondary packaging, and into the steps inside each stage: raw material verification, PPE, area preparation and release, content handling, storage.
• Release and cleaning checklists — including clean-in-place, which gate whether a stage may start at all.
• The occurrence and deviation form — where an operator records what departed from the procedure, at the moment it departed.
• In-process controls — the checks that carry a parameter, a limit and a deadline, and that decide whether the batch is still within specification.

They are one workflow in practice: a checklist gates a stage, the instruction runs it, an in-process control measures it, and an occurrence records where it went sideways. Digitising them separately reproduces the coordination cost that made paper expensive.

A paper record cannot tell an operator that a step is out of sequence, that a value is outside its limit, or that a required entry was skipped. Every one of those checks is deferred to a reviewer reading the record after the batch exists — which is why review is exhaustive and why a deviation found there becomes an investigation rather than a prompt.

The cost is not only the checking. A paper batch record passes through five hands before release: a supervisor prints it, operators fill it in, supervision follows the process, an assistant or coordinator checks the entries, and quality approves or rejects. Each handover is a queue, and the record spends far more time waiting than being worked on.

Moving the control to execution time changes four things at once:

• Sequence — steps run in the order the approved procedure defines, and skipping is not available rather than discouraged.
• Limits — an out-of-range entry is caught while the operator is still standing at the equipment, when correcting it is cheap.
• Second-person verification — requested where the procedure requires it, from someone who did not perform the step, and recorded as an attributable act.
• Completeness — a record cannot be closed with a required field empty, so the gap is never discovered weeks later.

That is also the precondition for review by exception: reviewers can look only at what deviated because the system classified the rest, and that classification is only trustworthy if the limits were under change control and the values were not transcribed.

This is the design decision that most affects what an EBR is worth, and the one most often left until last. If in-process values are typed by an operator reading a display, the electronic record inherits the weaknesses of the paper one: the timestamp is the typing time, the value is a transcription, and a second person is needed precisely because of that.

When the in-process control reads the measurement from NEO PI, the value arrives with the timestamp the device emitted and the quality code the protocol carried, traceable back to the raw event. The operator confirms instead of transcribing, and a whole class of error — and of verification — stops existing.

A third class of value is neither typed nor read from a device but calculated. NEO EBR lets you define the formulas and equations behind an input — yield, corrected weight, concentration, a potency-adjusted quantity — so the record computes them automatically from the values it already holds, the same way every time, timestamped and traceable to its operands. The manual arithmetic error, and the second person re-checking the math, stop existing too.

The same reasoning applies to material identity and quantity, which is why NEO W&D usually follows as the second record, and to unit-level identifiers, where NEO TRACK links what was shipped back to how it was made.

The batch record carries more regulatory weight than any other document a plant produces, so the properties that make it defensible are architectural rather than procedural.

1. Every action is attributable to an individual account, with electronic signature features designed to support 21 CFR Part 11 requirements — a signing PIN distinct from the login password, with its meaning declared at the moment of signing.
2. Configuration and data changes are recorded in a hash-chained audit trail with integrity verification, so retroactive alteration is detectable rather than merely prohibited.
3. Process values referenced from the historian keep their origin timestamp, which is what makes the record contemporaneous in the sense ALCOA+ means.
4. Master records are versioned under change control, so a batch can be shown against the exact procedure in force when it ran.

Qualifying the system for your regulatory context remains your validation activity — our note on computerised system validation under GAMP 5 covers keeping that effort proportional to risk. Both sit inside a wider digital compliance posture, and the paper-to-paperless roadmap sets out where the batch record belongs in the sequence.