The Overlooked Station Metric: Hand-Arm Vibration Is Showing Up in Supplier Audits

The acceptance sheet for a screw removal station usually carries four rows: cycle time, torque accuracy, first-pass yield and running stability. Rarely is there a row asking how much vibration an operator's wrist absorbs over a shift. Yet in 3C contract manufacturing supplier audits, that row is appearing, and the question is usually not whether the value is compliant but how the plant can prove it is.

Start with the measurement basis. Hand-arm vibration is expressed as the frequency-weighted acceleration at the hand grip, written ahv, computed from the three axes per ISO 5349 and then energy-averaged over the actual exposure duration to give an eight-hour equivalent value A(8). China's GBZ 2.2 sets an occupational exposure limit of 5.0 m/s² for a four-hour energy-equivalent weighted acceleration. EU Directive 2002/44/EC draws two lines: an exposure action value of 2.5 m/s² and an exposure limit value of 5.0 m/s², with risk assessment and control measures required once the action value is reached. The numbers are easy to look up. The hard part is applying them to one specific removal station.

Micro-screw stations are in fact more likely to cross the line, and the reason lies in the process itself. The smaller the screw, the finer the bit, and to keep a stable engagement on a limited head area the operator unconsciously increases downward pressure and grip force. Grip force amplifies vibration transmission: the harder the grip, the more energy reaches the arm. On heads between M0.8 and M1.2, a slight bit runout causes cam-out, so the operator repeatedly repositions and taps the trigger, producing far more trigger events per workpiece than a station handling M3 and above. Many short trigger events feel less tiring, but A(8) accumulates energy, so brief dense exposure is not safer than one continuous long exposure. Add the high spindle speeds typical of micro-screw drivers, and the number of vibration cycles per unit time rises further.

Station architecture matters more than the tool. A purely handheld station is the worst case: the full tool weight and reaction torque rest on the wrist. A balancer or torque-arm station transfers weight to the balancer, lowering grip force and exposure, though reaction torque remains. A desktop or vertical automated station leaves the operator with loading, confirmation and unloading only, and exposure approaches zero. This difference seldom enters an ROI model. Energy savings, headcount and yield all get calculated; the removal of an occupational health risk line usually stays in the EHS records and never reaches the capital justification sheet. For a plant that has to pass a customer EHS audit, however, that value is real.

The health benefit of automation only becomes meaningful when it maps to a specific structure. On a desktop small screw removal equipment platform, for example, the operator's task collapses into load, start and unload. The axial pressing force that used to come from the wrist is now supplied by a Z-axis cylinder or servo, and reaction torque is absorbed by the frame. This does not transfer risk to the machine; it removes an exposure that otherwise accumulates over time. A simple comparative metric works for evaluation: before the change it is grip minutes per shift multiplied by tool vibration value; after the change it is load cycles per shift multiplied by confirmation time. The latter is typically an order of magnitude smaller.

Field measurement is the step most often done wrong. ISO 5349-2 requires a triaxial accelerometer mounted close to the grip position and measured under real working conditions, not free running. Three common errors: measuring only free running, which usually understates the value; measuring a single axis and ignoring the difference between axial and radial; and substituting the longest continuous working period for the energy averaging in A(8), which turns intermittent work into continuous work. Where measurement is not feasible, a conservative estimate from the tool's declared vibration value plus exposure time is acceptable, provided the record states clearly that it is an estimate rather than a measurement. That survives scrutiny far better than a number of unknown origin.

Vision-equipped stations bring an additional benefit here. Human alignment relies on eyesight and feel, and feel is exactly where grip force comes from. Once alignment is handed to a camera and motion control, the operator shifts from continuous gripping and fine adjustment to placing and confirming, and grip time drops sharply. An automatic vision screw removal machine is increasingly common in rework and disassembly, partly for accuracy and consistency, and partly for a plainer reason: it takes the operator out of a state where the tool must be held continuously.

What auditors pursue is the evidence chain rather than a single figure. A defensible package typically contains a station list with actual exposure duration per shift, tool model with declared or measured vibration value, the A(8) calculation, and the control measures already in place with their priority. The priority order is settled: engineering controls before administrative controls, administrative controls before personal protective equipment. Anti-vibration gloves offer limited attenuation, almost none at low frequencies, and the extra thickness forces a stronger grip that can cancel what little attenuation exists. Presenting gloves as the primary measure is therefore rarely accepted in a professional audit.

For buyers, the practical conclusion is to treat vibration exposure as a design input alongside cycle time and yield. It can be written into a requirement document in four statements: how long the station is gripped per shift; what the tool's hand-arm vibration value is; whether A(8) stays below 2.5 m/s²; and if not, which engineering control brings it down. The first three are assessment; only the fourth is specification. Once the fourth appears, suppliers tend to move from offering a lower-vibration driver to asking whether a human should be holding the tool at all, which is the more useful question.

In the M0.8 to M2.0 range this question is especially valid. The smaller the screw, the longer the handheld exposure, the higher the cam-out probability and the more trigger events per part. With all three stacking, mechanisation pays back faster than at larger screw stations. An M0.8-M2.0 screw removal solution normally lists manual grip time as a separate input during project evaluation for exactly this reason: it affects occupational health compliance, yield stability and long-term labour cost at once, one of the few parameters that pays back in three places from a single change.