Case study — recycling operator replaces belt-scale logic with volume flowCase study — recycling operator replaces belt-scale logic with volume flowCase study — recycling operator replaces belt-scale logic with volume flow

The starting point. A north-German recycling operator processes mixed input material — wood, plastic, metal mix from commercial waste — through a large shredder. On the feed belt sat a belt scale rated by the manufacturer at 0.5 % accuracy. In practice: tonnage reading drifted by 4–6 % per week depending on material mix and belt soiling, calibration shifts every 6–8 weeks, recurring discussions with customers about the billed quantity.

Why the belt scale did not work. Not because it was bad, but because the precondition was missing: stable bulk density. In a typical recycling mix the specific weight between layers swings by a factor of 2–3. A belt scale measures mass correctly, but the downstream balance with the goods-in records still does not close, because material is lost between load cell and shredder (at the transfer chute, in spillage). That only becomes visible from tonnage-per-hour curves which, without a continuous volume profile, no one understands. On top of that: material build-up on the weigh idlers slowly drifts the reading, and the quarterly calibration shifts cost 6–8 hours of full-line downtime each.

The installation. In February 2023, OWL EYE® VOLUME FLOW was installed on the feed belt to the shredder, with an OWL EYE® STOCKPILE in parallel on the pre-pile from which material flows onto the belt. Sensor cluster on a cross-belt mast, IP66 housing, air-purge optic cleaning against wood and plastic dust. Data over OPC-UA into the existing SCADA.

What has actually changed — after 14 months in operation.

Position	Before (belt scale)	After (VOLUME FLOW + STOCKPILE)
Tonnage accuracy	4–6 % drift/week	±1.5 % continuous
Calibration intervals	every 6–8 weeks	1× at commissioning + annual check
Downtime for calibration	6–8 h per shift, 6–8×/year	0 h
Data history	daily shift totals	minute-resolved volume + tonnage
Personnel effort belt scale	~80 h/year supporting calibration	~8 h/year sensor visual check
Maintenance window	permanently scheduled	no longer needed

The belt-scale story. The belt scale was not removed — it stays in place for trade-relevant weighing at the dispatch gate, and it continues to deliver a second measurement value for plausibility checks. What disappeared is the constant calibration shifts and the customer arguments — the LiDAR volume data from the feed is now the primary control signal, and the belt scale is the billing-relevant secondary measurement.

ROI summary. From eliminated calibration shifts (8 shifts × 7 h × roughly €4,000 of line-stop cost) plus reduced personnel effort on belt-scale service, gross savings worked out at about €35,000/year. The investment in the sensor cluster plus engineering was in the low six figures. Payback: about 14 months.

What the operator says in hindsight. The real win is not the ROI number, but the disappearance of the constant "the belt scale shows something different than the delivery note" debate. With two independent measurements — volume from LiDAR plus mass from the belt scale — discrepancies can immediately be attributed to a source (material drifts vs. scale drifts) instead of triggering another calibration shift to find out.

What this is not. This case study is not a universal proof that LiDAR replaces a belt scale. It is one example that, in a recycling mix with strongly varying bulk density, a continuous volume signal noticeably eases operational control — and that the investment pays back from eliminated downtime, not from "better tonnage values".

More on volume measurement for recycling at /recycling/, plus /volume-flow/, /industries/ and in our background article LiDAR vs. belt scale.

The starting point. A north-German recycling operator processes mixed input material — wood, plastic, metal mix from commercial waste — through a large shredder. On the feed belt sat a belt scale rated by the manufacturer at 0.5 % accuracy. In practice: tonnage reading drifted by 4–6 % per week depending on material mix and belt soiling, calibration shifts every 6–8 weeks, recurring discussions with customers about the billed quantity.

Why the belt scale did not work. Not because it was bad, but because the precondition was missing: stable bulk density. In a typical recycling mix the specific weight between layers swings by a factor of 2–3. A belt scale measures mass correctly, but the downstream balance with the goods-in records still does not close, because material is lost between load cell and shredder (at the transfer chute, in spillage). That only becomes visible from tonnage-per-hour curves which, without a continuous volume profile, no one understands. On top of that: material build-up on the weigh idlers slowly drifts the reading, and the quarterly calibration shifts cost 6–8 hours of full-line downtime each.

The installation. In February 2023, OWL EYE® VOLUME FLOW was installed on the feed belt to the shredder, with an OWL EYE® STOCKPILE in parallel on the pre-pile from which material flows onto the belt. Sensor cluster on a cross-belt mast, IP66 housing, air-purge optic cleaning against wood and plastic dust. Data over OPC-UA into the existing SCADA.

What has actually changed — after 14 months in operation.

Position	Before (belt scale)	After (VOLUME FLOW + STOCKPILE)
Tonnage accuracy	4–6 % drift/week	±1.5 % continuous
Calibration intervals	every 6–8 weeks	1× at commissioning + annual check
Downtime for calibration	6–8 h per shift, 6–8×/year	0 h
Data history	daily shift totals	minute-resolved volume + tonnage
Personnel effort belt scale	~80 h/year supporting calibration	~8 h/year sensor visual check
Maintenance window	permanently scheduled	no longer needed

The belt-scale story. The belt scale was not removed — it stays in place for trade-relevant weighing at the dispatch gate, and it continues to deliver a second measurement value for plausibility checks. What disappeared is the constant calibration shifts and the customer arguments — the LiDAR volume data from the feed is now the primary control signal, and the belt scale is the billing-relevant secondary measurement.

ROI summary. From eliminated calibration shifts (8 shifts × 7 h × roughly €4,000 of line-stop cost) plus reduced personnel effort on belt-scale service, gross savings worked out at about €35,000/year. The investment in the sensor cluster plus engineering was in the low six figures. Payback: about 14 months.

What the operator says in hindsight. The real win is not the ROI number, but the disappearance of the constant "the belt scale shows something different than the delivery note" debate. With two independent measurements — volume from LiDAR plus mass from the belt scale — discrepancies can immediately be attributed to a source (material drifts vs. scale drifts) instead of triggering another calibration shift to find out.

What this is not. This case study is not a universal proof that LiDAR replaces a belt scale. It is one example that, in a recycling mix with strongly varying bulk density, a continuous volume signal noticeably eases operational control — and that the investment pays back from eliminated downtime, not from "better tonnage values".

More on volume measurement for recycling at /recycling/, plus /volume-flow/, /industries/ and in our background article LiDAR vs. belt scale.

The starting point. A north-German recycling operator processes mixed input material — wood, plastic, metal mix from commercial waste — through a large shredder. On the feed belt sat a belt scale rated by the manufacturer at 0.5 % accuracy. In practice: tonnage reading drifted by 4–6 % per week depending on material mix and belt soiling, calibration shifts every 6–8 weeks, recurring discussions with customers about the billed quantity.

Why the belt scale did not work. Not because it was bad, but because the precondition was missing: stable bulk density. In a typical recycling mix the specific weight between layers swings by a factor of 2–3. A belt scale measures mass correctly, but the downstream balance with the goods-in records still does not close, because material is lost between load cell and shredder (at the transfer chute, in spillage). That only becomes visible from tonnage-per-hour curves which, without a continuous volume profile, no one understands. On top of that: material build-up on the weigh idlers slowly drifts the reading, and the quarterly calibration shifts cost 6–8 hours of full-line downtime each.

The installation. In February 2023, OWL EYE® VOLUME FLOW was installed on the feed belt to the shredder, with an OWL EYE® STOCKPILE in parallel on the pre-pile from which material flows onto the belt. Sensor cluster on a cross-belt mast, IP66 housing, air-purge optic cleaning against wood and plastic dust. Data over OPC-UA into the existing SCADA.

What has actually changed — after 14 months in operation.

Position	Before (belt scale)	After (VOLUME FLOW + STOCKPILE)
Tonnage accuracy	4–6 % drift/week	±1.5 % continuous
Calibration intervals	every 6–8 weeks	1× at commissioning + annual check
Downtime for calibration	6–8 h per shift, 6–8×/year	0 h
Data history	daily shift totals	minute-resolved volume + tonnage
Personnel effort belt scale	~80 h/year supporting calibration	~8 h/year sensor visual check
Maintenance window	permanently scheduled	no longer needed

The belt-scale story. The belt scale was not removed — it stays in place for trade-relevant weighing at the dispatch gate, and it continues to deliver a second measurement value for plausibility checks. What disappeared is the constant calibration shifts and the customer arguments — the LiDAR volume data from the feed is now the primary control signal, and the belt scale is the billing-relevant secondary measurement.

ROI summary. From eliminated calibration shifts (8 shifts × 7 h × roughly €4,000 of line-stop cost) plus reduced personnel effort on belt-scale service, gross savings worked out at about €35,000/year. The investment in the sensor cluster plus engineering was in the low six figures. Payback: about 14 months.

What the operator says in hindsight. The real win is not the ROI number, but the disappearance of the constant "the belt scale shows something different than the delivery note" debate. With two independent measurements — volume from LiDAR plus mass from the belt scale — discrepancies can immediately be attributed to a source (material drifts vs. scale drifts) instead of triggering another calibration shift to find out.

What this is not. This case study is not a universal proof that LiDAR replaces a belt scale. It is one example that, in a recycling mix with strongly varying bulk density, a continuous volume signal noticeably eases operational control — and that the investment pays back from eliminated downtime, not from "better tonnage values".

More on volume measurement for recycling at /recycling/, plus /volume-flow/, /industries/ and in our background article LiDAR vs. belt scale.