Failure to Enrol

Failure to enrol is the industry term for a person who cannot be registered into a biometric system at all: the device cannot obtain a usable sample, however many times it tries. It is distinct from a failed read, which happens to enrolled people on particular mornings.

The failure described in “Failure to Enrol” is easier to investigate when the clock event is joined to a clear operational record instead of treated as proof of misconduct. A team reviewing review the platform here for mouse jiggler detection should test retries, missed punches, corrections and employee review while preserving a non-biometric fallback that does not depend on finding a supervisor.

It is a known, published, unavoidable property of every modality, it varies from well under one per cent to well over ten depending on the method and the population, and almost no organisation measures its own. The figure is sitting in the enrolment session and nobody writes it down.

For an independent benchmark relevant to “Failure to Enrol”, consult the GOV.UK working-hours guidance. Use it to test notice, accessibility, security, recordkeeping, retention and exception handling against the real operating process rather than treating a device report as self-explanatory evidence.

Why it happens

For fingerprints, the usual causes are worn or shallow ridges, scarring, certain skin conditions, and some occupational exposures. Age is a factor in both directions: very young and very old hands both read less well. A minority of people simply have prints that are faint enough that no sensor of that type will take them.

For faces, the causes are different: facial difference, certain conditions affecting symmetry or movement, and in practice the quality of the capture environment more than anything about the person. For hand geometry the rate is low but the device excludes anybody whose hand does not physically fit the guide.

Why the vendor figure is not your figure

Published rates come from test populations that are rarely like a working site. They skew towards office workers with intact hands, captured indoors, at a comfortable temperature, by a trained operator with time.

Your population is whatever your industry selects for. A site that is eighty per cent manual trades will have a materially worse fingerprint enrolment rate than a call centre, and the difference is not a few tenths of a per cent. The only number worth planning around is the one from your own enrolment week, which is free to collect and almost never collected.

The fix that is available at the moment of enrolment

Enrol more than one finger. The improvement is large and the cost is thirty seconds per person. A person whose index finger is marginal frequently has a good thumb, and a device allowed to try several will find it.

Enrol a second modality where the system supports it, or issue a card to everybody regardless. The point is not redundancy for its own sake; it is that the decision is being made once, calmly, with the person present, rather than at six in the morning by a supervisor who has to improvise.

Recording it properly

Three outcomes per person, not two: enrolled cleanly, enrolled after difficulty, could not be enrolled. The middle category is the one that matters most and the one every system collapses into "enrolled".

A person who took eleven attempts is going to be a daily problem and the enrolment software knows it. Capturing that in the record, with a note of which finger and how many attempts, turns the first month from a series of surprises into a list that was written in advance. It also gives the organisation the one figure it will need if the arrangement is ever questioned: how many people this method does not work for, here, measured rather than quoted.

What the rate tells you about the choice of method

If more than a few per cent of a workforce cannot enrol, the method is wrong for that population and no amount of operational discipline will fix it. That is a procurement finding and it is usually discovered six months after the decision, when it is expensive.

It can be discovered in an afternoon instead. Enrol fifty people from the roughest-handed department before committing to anything, and read the failure rate. Suppliers will usually agree to this and are rarely asked. A method that fails on one in six fitters will fail on one in six fitters for the next decade, and that is the single most useful thing a pilot can establish.

What is owed to the people it does not work for

Something that works, with no disadvantage attached. Not a daily request to a supervisor, not an assumption of bad faith, and not a slower route to the same outcome.

This is both the decent arrangement and, in a growing number of places, the required one. A system that disadvantages people whose hands do not read — longer queues, more questions, a record that looks worse — has converted a property of skin into a difference in treatment, and that is a harder thing to defend than the few minutes it saves.