Vehicle entry, weighment and inward quantity are recorded on paper at the gate, so the goods receipt cannot be reconciled against what physically arrived, and gate queues are invisible to the plant.

The gate reads the number plate, captures weighbridge readings automatically and logs the inward entry against the vehicle, so the receipt is built from measured data rather than from a register.

Machine utilisation is estimated from memory at the end of the shift. Nobody can say how long a machine actually ran, how often it stopped, or what the real availability was.

A monitor on the machine counts output against a pro-rated target, records every stop with its reason, and computes availability, performance and quality into a live OEE figure on the machine and in the browser.

Critical process parameters are checked on rounds and written into a logbook. Drift between rounds is discovered only after it has affected the batch.

Parameters are logged continuously on multiple channels with high and low alert bands, so a deviation raises an alert while it can still be corrected, and the batch record writes itself.

Inspection results live on paper checklists. Rejections are counted at the end of the shift, and by then it is impossible to tie them back to a line, a shift or an operator.

Rejections are counted as they happen alongside production, and inspection is captured against a checklist so that quality is reported per line and per shift rather than as one monthly number.

A short carton reaches the customer before anyone notices. Pack counts and rejects are reconciled by hand at the end of the shift.

Vision at the packing station verifies the contents of every carton before it is sealed and counts packs automatically, so a short or wrong pack is stopped at the line instead of at the customer.

Boiler and temperature readings are entered into a register by hand for the auditor. An excursion overnight is discovered the next morning, if at all.

Temperatures are logged continuously with alarm relays and e-mail alerts on excursion, giving both the live warning and the audit record from the same instrument.

One electricity bill arrives for the whole plant. There is no way to know which line, section or utility is consuming what, so there is no way to argue with the number.

Energy is metered section by section and logged continuously, turning a single bill into a consumption profile per line, per shift and per unit produced.

Maintenance happens after the breakdown. Service intervals are based on the calendar rather than on how many hours the machine actually ran.

Run hours and stoppages are captured from the machine itself, so servicing is scheduled against real running time and repeat failures surface as a ranked list rather than as anecdote.

Cathodic protection on a buried pipeline is verified by sending a technician to each test post with a handheld meter. Readings are sparse, manually recorded, and a protection failure between surveys goes unnoticed until corrosion has already progressed.

Remote CP loggers record pipe-to-soil potential continuously at each test post and report to the cloud, while synchronised current interrupters allow IR-free on/off potentials to be captured across the survey without a technician at every station.

Attendance comes off two or three biometric machines in different formats and is reconciled into payroll by hand every month. Leave lives on paper, so the salary sheet and the leave register rarely agree.

Punches from every machine are imported and merged per employee per day, leave runs through an approval workflow, and salary slips are computed from the same attendance record that HR already verified.

Appraisals are written from whatever the reviewer remembers of the last three months, which makes the rating hard to defend and harder for the employee to accept.

Performance is tracked against defined KPIs through the year and combined with attendance analytics, so the appraisal conversation starts from a record instead of from recollection.

Enquiries sit in individual notebooks and inboxes. Nobody can say how many are open, which are worth chasing, or which were never followed up at all.

Every enquiry is captured, scored and moved through a defined pipeline, so follow-up is a process with visible status rather than a matter of who remembers.