Side-by-side parcel detection refers to the condition in which two or more discrete parcels enter an induction aperture or metering belt with insufficient longitudinal separation, such that the system cannot resolve them as independent units. This condition is distinct from a fully jammed or stalled line, because the conveyor continues to move, the parcels appear to be moving normally, and the induction system may still be cycling. However, the downstream singulator, camera array, or dimensioning device cannot determine where one parcel ends and the next begins. For courier hubs, side-by-side events are a leading contributor to mis-sorts, missed dimensioning charges, and unexpected mechanical stops at the merge point. This article describes the physical zones where side-by-side conditions originate, the observation points that give the earliest warning, and the practical boundaries that maintenance and controls teams should apply when responding.
Induction Context and Parcel Flow Dynamics #
Induction in a parcel depot typically begins at a receiving lane or unload dock where bulk parcels are tipped onto a flat conveyor. From there, parcels pass through a series of orientation and spacing devices before being presented to a dimensioning scanner and ultimately a singulator or merge section that releases parcels onto the main sortation loop. A side-by-side condition is most likely to develop in the short window between the bulk receiving conveyor and the metering belt, where parcels are still transitioning from a dense, randomized flow to a spaced, single-file presentation.
At a courier hub, parcel lengths can vary from small poly bags to oversized cartons. The induction system must therefore make a continuous decision about parcel boundaries. When two parcels stand side by side on a single belt, they present a combined silhouette that may appear to the controls system as one very wide or unusually shaped item. In some systems, the control logic will attempt to treat this combined silhouette as a single parcel, which causes the downstream sorter to receive an inaccurate length, width, or height value. In other systems, the controller will see a discontinuity in the edge profile and attempt to split the silhouette, producing two partial parcel records that fail to match the physical units being conveyed.
The operating context is important because side-by-side events are not always a mechanical failure. They can occur during peak surge when unload teams place parcels onto the belt faster than the singulator can recover. They can also occur when a parcel is tilted, leaning on its neighbor, or resting on top of a second parcel in a way that creates a false side-by-side profile. The detection challenge is therefore both physical and logical: the system must identify that a boundary violation exists while also determining whether the perceived boundary is the result of a true second parcel, a shadow, a void, or a strap or label artifact.
Why Side-by-Side Parcels Stress the Downstream System #
When a side-by-side parcel pair passes undetected past the induction metering belt, the immediate consequence is often a misread at the dimensioning or identification zone. The dimensioning camera will measure the combined footprint, producing a volumetric data point that is too large for both parcels and too small for a single consolidated package. That erroneous measurement propagates to the sortation controller, which then assigns a destination lane based on false data. In many hubs, this results in the parcel pair reaching a chute that is not wide enough for the combined footprint, causing a jam at the chute entry.
The second major effect is on the singulator or merge wheel. A side-by-side pair that enters the singulator with only a narrow gap may be treated as one wide parcel. The singulator will attempt to guide that wide item into a single lane, which can cause the leading parcel to be skewed, pushed ahead, or rotated. If one parcel is heavier than the other, the lighter parcel may be dragged along or flicked off the belt edge. In extreme cases, the heavier parcel will slow down while the lighter parcel accelerates, creating a dynamic gap that is not predictable to the photo-eye array.
From a controls perspective, side-by-side events also create a synchronization problem. The system tracks parcels by their leading edge position and expected speed. When a side-by-side pair is later split by the singulator, the two resulting parcels will appear to the system as if they were spawned from nothing, or as if a parcel has duplicated. This causes the tracking buffer to lose alignment, which may cascade into false trigger events at the sortation induction point. Maintenance teams often observe the downstream symptoms, such as unexpected empty sortation cells or repeated jams at the same merge point, without immediately realizing that the root cause was the earlier side-by-side condition at induction.
Primary Inspection Points #
The earliest warning of a side-by-side problem is not found at the dimensioning scanner, where the failure becomes visible, but rather at the upstream zones where the condition is created. A structured inspection should cover at least four zones: the bulk receiving belt, the singulator feed section, the metering belt, and the dimensioning or scanning void.
Zone 1: Bulk Receiving Belt and Unload Lane #
Inspect the bulk belt for accumulation patterns. Look for places where parcels are stacked two-high along the belt edges, or where the unload team has placed long parcels perpendicular to the direction of travel. A parcel that is slightly angled across the belt will reduce the effective width available for neighboring parcels, increasing the chance that the next parcel slides alongside it rather than behind it. Also check whether the belt is running at the same speed as the upstream discharge. If the receiving belt is slower than the unload conveyor, parcels will tend to bank up and slide sideways rather than forming a clean queue.
Another inspection point in this zone is the vertical clearance above the belt. Parcels that pile upward can form a secondary layer that falls forward onto the belt, creating a transient side-by-side profile with a parcel that is already being conveyed. This is especially common with poly bags and lightweight items that compact easily. The unload side should show clear flow lanes with no persistent pile height above approximately one parcel depth.
Zone 2: Singulator Feed Section #
The singulator feed section is the transition where the wide bulk flow narrows into one or more lanes. This is the most critical inspection point for side-by-side events. Verify that the guide rails or fence walls are parallel and free of distortion. A fence that has been bent outward by a heavy parcel will create a wider channel, allowing two average-size parcels to travel side by side. Check that the belt height is consistent at the transition. A sagging belt section can create a low point where parcels drift together, eliminating the small gap that otherwise separates them.
Listen for unusual belt slip or rattling in this area. If the singulator feed belt is losing traction, it may intermittently slow down while the upstream belt continues at full speed. The resulting compression causes parcels to slide sideways into neighboring flow lanes. The controls system may not register this as a fault because the belt motor is still running and the photo-eye state is normal.
Zone 3: Metering Belt and Staging Lanes #
On the metering belt, inspect the side guide spacing and the condition of the belt surface. A worn rubber surface with poor friction can cause parcels to skid when the metering belt starts and stops. When a metering belt cycles rapidly, the rear parcel may continue sliding forward during the stop phase, closing the gap to the parcel ahead and eventually overlapping it. The result is a side-by-side or tandem overlap that is difficult for the downstream scanner to separate.
Also inspect the staging lane length. If the staging area is too short for the longest parcels being inducted, the rear of a long parcel may still be on the upstream belt when the front enters the metering belt. This can cause the parcel to rotate and present its side to the next parcel, creating a false wide profile. Check that the staging lane length matches the maximum parcel length that the hub is configured to process.
Zone 4: Dimensioning and Scanning Void #
The dimensioning void is where the side-by-side condition becomes measurable. Here, the inspection focus is on the camera field of view and the background surface. Look for reflective or dark patches on the belt that may confuse the edge detection algorithm. A dark patch on a light belt can be interpreted as a gap between parcels, while a reflective patch may cause the camera to miss the actual gap. Also inspect the lighting array for flicker or uneven intensity, as this can affect the edge contrast that the detection algorithm relies on.
Position a test object, such as a cube or rectangular box of known dimensions, at the location where side-by-side parcels typically appear. Run the test object through the dimensioning zone at a representative speed and compare the measured footprint to the known dimensions. If the measurement is inconsistent at one side of the belt, the camera alignment or lighting is likely the contributing factor.
Observable Symptoms and Diagnostic Table #
The following table outlines the common symptoms of side-by-side parcel detection issues, the likely zone of origin, and the immediate evidence to collect. This table is a practical starting point for a root-cause discussion, not a definitive fault list. Actual conditions will vary by conveyor manufacturer and system configuration.
| Observed Symptom | Likely Origin Zone | Evidence to Collect | Early Warning Character |
|---|---|---|---|
| Frequent jams at the merge wheel or singulator exit | Zone 2: singulator feed | Photo-eye timing logs showing short or inconsistent gaps between parcels | Jam occurs only when two mid-size parcels enter the singulator within 300 ms of each other |
| Dimensioning scanner reports oversized width on random parcels | Zone 4: dimensioning void | Camera snapshot of the parcel footprint showing two distinct edge shadows | Oversized width measurement repeats for multiple parcels with no actual physical size change |
| Sortation controller shows empty cells or phantom parcels | Zone 3: metering belt | Tracking buffer log showing a parcel record created without a matching leading-edge trigger | Empty cells occur downstream only after a period of high induction rate |
| Parcels with scuff marks on the leading or trailing edge | Zone 1: bulk receiving belt | Photographs of the damaged parcel edges and the corresponding belt section | Marks appear consistently on parcels originating from one side of the receiving belt |
| Recurring mis-sorts to the same destination lane | Zone 4: dimensioning void or Zone 2 | Sortation data showing that mis-sorted parcels share a similar length-to-width ratio | Mis-sort rate increases when parcel throughput rises above a specific threshold |
| Audible thumping from the metering belt during high induction | Zone 3: metering belt | Vibration or accelerometer reading showing a periodic pulse synchronized with parcel release | Thumping occurs at the same belt cycle every time and is not present at low throughput |
Evidence Collection and Data Logging #
Effective diagnosis of side-by-side events depends on collecting the right evidence at the right time. Begin with the control system event log. Identify the timestamps when the dimensioning scanner flagged an oversized width or when the sortation controller recorded a missed trigger. Then correlate those timestamps with the belt speed and the induction photo-eye states. In many depots, the controls platform already records this data, but it is not commonly reviewed in a time-aligned view. Export the relevant window and align the traces on a single timeline.
Next, collect photographic evidence from the fixed cameras at the induction area. If the hub has a camera overlooking the bulk receiving belt, retrieve the video footage from the identified time window. Look for the ten seconds before the side-by-side event, not just the moment of the event. The behavior that leads to a side-by-side condition often occurs in the seconds prior, when a parcel is placed too close to another or when a small parcel is caught in the wake of a large one.
Where automated data is not available, perform a manual observation. Position a person at the Zone 2 singulator feed section for a full induction cycle, or a minimum of fifteen minutes, and record the instances where two parcels travel with no visible gap. Note the belt speed, the approximate parcel length, and the direction from which the parcels arrived. Repeat this observation at different times of day to capture the difference between a light, steady flow and a heavy, surging flow.
Also record the mechanical setup at the time of the event. Measure the gap between guide rails at several points in Zone 2. Compare that measurement to the nominal setting documented in the OEM manual. Document whether the belt has been recently replaced or retensioned, and whether any belt cleaner or scraper was adjusted. A measurement taken at the moment of the fault is far more useful than a measurement taken after the system has been cleared and reset.
Common Interpretation Errors #
One recurring error in the field is to attribute side-by-side events solely to operator speed at the unload lane. While unload pace does influence the condition, the system should be designed to tolerate short bursts of high throughput. If side-by-side conditions occur regularly at the same belt speed, the cause is more likely a mechanical or controls issue than operator behavior. It is important not to blame operator performance before confirming that the singulator feed can actually handle the rated throughput with the observed parcel mix.
A second interpretation error is to confuse side-by-side events with parcel skew. A skewed parcel occupies a diagonal footprint on the belt, which may appear wider to the dimensioning camera. However, a skewed parcel typically has a measurable leading and trailing edge, allowing the tracking system to infer the actual orientation. A side-by-side pair has two leading edges and two trailing edges, but the gap between them is below the detection threshold. These are distinct failure modes and they require different corrective actions. Skew is corrected by adjusting the fence alignment or the belt tracking; side-by-side is corrected by increasing the gap between parcels or by changing the physical spacing mechanism.
A third error is to assume that the dimensioning scanner is the most reliable detection point. In many induction systems, the dimensioning scanner is located too close to the singulator to provide meaningful early warning. By the time the scanner sees the side-by-side condition, the parcels are already past the point where a control action can separate them. The reliable detection point is upstream, at the metering belt or the singulator feed, where a photo-eye or sensor array can measure the gap between parcel edges and trigger a corrective action. The dimensioning scanner should be used for confirmation, not for primary detection.
A fourth error is to treat all side-by-side events as a single problem. The condition can be a permanent mechanical issue, such as a bent fence rail, or a transient controls issue, such as a missed photo-eye trigger that leaves a parcel in the staging lane. These two cases require very different responses. A permanent issue is best addressed by a mechanical correction. A transient issue is best addressed by reviewing the photo-eye timing and the control logic that determines when the metering belt releases a parcel. The maintenance team should separate the evidence into mechanical and logical categories before deciding on a fix.
Maintenance Implications #
Preventive maintenance should include a routine check of the guide rails and fences in the singulator feed zone. Measure the rail-to-belt clearance and the rail-to-rail spacing at least weekly during peak season, and record the measurements so that drift can be identified early. A rail that moves outward by only five millimeters can be enough to allow a side-by-side condition for the most common parcel widths in a courier network.
Belt condition is another high-impact maintenance item. A belt with uneven wear or localized surface damage will have variable friction across its width. This causes one side of a parcel to be pulled faster than the other, which can rotate a parcel and encourage side-by-side contact with the neighboring item. Inspect the belt longitudinally for worn areas that correspond to the position where parcels most frequently travel. Replace the belt if the surface profile is no longer flat, rather than attempting to compensate with upper hold-down rollers.
The photo-eye array at the metering belt should be cleaned and tested at regular intervals. Dust and strap debris can accumulate on the lens surface and reduce the sensitivity to low-contrast parcel edges. A photo-eye that is marginal in sensitivity may detect the gap between a light parcel and a dark conveyor belt, but fail to detect the gap between a light parcel and a light-colored poly bag. Run a periodic test using a set of low-contrast test parcels to confirm that the sensor is responding correctly.
Finally, include the side-by-side condition in the shift handover checklist. The controls team should note any unusual patterns in the tracking buffer, and the mechanical team should note any adjustments made to the singulator feed. The two teams share responsibility for this condition, and a handover form that captures both the logical and the mechanical observations will reduce the time to root cause on the next occurrence.
Decision Boundaries #
Establishing a decision boundary for side-by-side events means defining when the condition requires immediate stop, when it can be handled by a temporary control adjustment, and when it is acceptable for the system to continue operating until the next planned maintenance window.
An immediate stop is warranted when side-by-side parcels are causing repeated jams at a point that requires a person to reach into the conveyor path, when the packaging of any parcel is being damaged, or when the tracking buffer has lost synchronization to the point that the sortation controller cannot reliably assign destinations. In these situations, the priority is safe system shutdown and clearance of the affected zone. Ensure that all site lockout and energy-isolation procedures are followed before anyone approaches the obstructed area.
A temporary control adjustment may be appropriate when the side-by-side events are caused by a marginal gap setting on the metering belt. The controls engineer may increase the release interval by a small increment, such as a few hundred milliseconds, to allow the singulator more time to recover. This adjustment reduces throughput but preserves system reliability. Document the change and share it with the operations team so that they understand the reduced induction rate. A temporary adjustment should not remain in place indefinitely; it should be followed by a root-cause investigation and a permanent correction at the next maintenance window.
Continuous operation is acceptable when the side-by-side event is rare, does not cause a mis-sort, and does not damage parcels. For example, a small poly bag that momentarily overlaps a larger carton but separates cleanly at the singulator may not produce any downstream effect. In this case, the system is operating within its designed tolerance, and forcing an unscheduled stop would create more disruption than the condition itself.
It is important to recognize that the decision boundary is not fixed for all lines in the same hub. A side-by-side tolerance that is acceptable on a low-speed induction lane feeding a manual sort area may be unacceptable on a high-speed induction lane feeding an automated cross-belt sorter. The decision boundary should be defined for each induction configuration by the maintenance and controls team, based on the specific downstream consequences that the hub is willing to accept.
Key Takeaways #
- Side-by-side parcel detection is a boundary condition, not a single point failure; it must be assessed across the bulk receiving belt, singulator feed, metering belt, and dimensioning void.
- The earliest reliable detection point is upstream, at the gap between the metering belt and the singulator, not at the dimensioning scanner.
- Distinguish side-by-side events from parcel skew; the two have different rotational mechanics and different corrective actions.
- Collect time-aligned evidence, including photo-eye logs, tracking buffer records, and video footage, before deciding on a mechanical or logical root cause.
- Guard against blaming operator pace when the physical flow path cannot support the rated throughput with the observed parcel mix.
- Review guide rail spacing and belt surface condition as part of routine preventive maintenance; these two items account for a large share of otherwise unexplained side-by-side conditions.
- Apply a staged decision boundary: immediate stop for jams or damage, temporary controls adjustment for marginal gap settings, and continuous operation for rare events that produce no measurable downstream effect.
- Site-specific safety procedures, lockout requirements, OEM documentation, and competent engineering judgment always take priority over any general guidance provided in this article.
