Device Grading Software for Refurbished Devices

A grade rule is a set of conditions over named inputs. Each input has a source, a threshold and a weight or blocking flag. The rule engine evaluates the inputs in a fixed order so two devices with the same evidence cannot receive different grades.

The order matters. A device with a cracked screen should not reach grade A because its battery is excellent, and a device with a failed camera should not be rescued by a clean body.

Rules should be written with the resale channel in mind. A device sold to a consumer needs a stricter cosmetic standard than one sold to a trade buyer who will refurbish it again. The same test readings can therefore produce different grades for different channels, as long as the channel is recorded.

Every rule needs a version. When a threshold changes, old grades should keep the rule version that produced them. Without that, a buyer comparing two reports from different months cannot tell whether the device changed or the standard did.

It also helps to define what the grade does not cover. A grade normally describes condition, not origin, warranty or legal status. Those belong in separate fields so a buyer does not read more into the letter than it promises.

Cosmetic inspection is the least standardised part of grading, so it needs structure. Instead of a free-text note, the technician marks regions on a device map, selects a mark type and severity, and adds a photo. The system counts the marks and applies the cosmetic band.

Functional results should never be flattened into a single pass or fail. A skipped test is not a failure, and a failed test is not the same as a component that is absent. The grading engine needs those states kept apart, because a device with an untested port is a different commercial risk from one with a broken port.

Photo evidence should be optional for the grade but mandatory for the report. A buyer who disputes a scratch wants to see the device as it was inspected, not a description written days later.

The grade has to travel with the device. A label should carry the asset or serial number, the grade, the date and a QR link to the report. The same fields should be available for a marketplace listing, so a reseller does not retype the condition.

The buyer report is a shorter version of the internal record. It shows the grade, the findings that affected it, the test summary and the cosmetic evidence, while hiding internal notes and cost data. Both versions come from the same record, which prevents the listing from drifting away from the bench result.

Where a marketplace has its own condition taxonomy, a mapping table translates the internal grade into the external one. The mapping is configuration, not code, because marketplaces change their categories more often than refurbishers change their standards.

It does not set the price. Price depends on demand, channel, age and margin, and those are commercial decisions that change faster than condition standards. The grade is an input to pricing, not a substitute for it.

It also does not certify origin or ownership. A stolen-device check, an IMEI status check and a warranty lookup are separate integrations, and a grade should never imply that those checks passed.