Pick-to-light systems are among the most predictable and controllable interfaces between human operators and order-fulfillment software. When properly sized, balanced, and maintained, they deliver consistent throughput with low cognitive load. When pushed past their design envelope, they produce deceptively hard-to-trace delays that appear as operator, software, or mechanical failures. This article provides a structured framework for capacity planning and bottleneck analysis for pick-to-light workstations, with emphasis on evidence collection, interpretation, and the boundaries of what an operations team can decide without OEM support.
Operating Context and the Role of Pick-to-Light #
Pick-to-light systems are most commonly deployed in zone-based piece-picking operations, batch-picking clusters, and goods-to-person workstation interfaces. The core concept is simple: a light module associated with a storage location or tote directs the operator to the next pick, displays the quantity, and confirms the action through a pushbutton, sensor, or barcode scan. In practice, however, the system is a tightly coupled chain of order release logic, zone controllers, network infrastructure, display hardware, mechanical fixtures, and human behavior. A capacity plan that only considers the fastest operator click rate will fail because it ignores the other links.
For warehouse operators, the practical value of understanding capacity planning is not merely maximizing picks per hour. It is the ability to predict when a workstation or zone will saturate, to identify which constraint is actually limiting flow, and to make informed decisions about rebalancing, automation, or expansion. For maintenance and controls teams, the same knowledge turns a vague “it feels slow” complaint into a testable hypothesis.
Components That Shape Capacity #
Every pick-to-light system has the same functional components, but the capacity characteristics of each component differ. The first step in planning or analysis is to build a mental model of these components and their interactions.
Light Modules and Display Units #
The display module is the visual interface. It shows the pick location, quantity, and sometimes additional text or color signals. Module response time is generally not a constraint in modern systems; however, the physical arrangement of modules—their height, viewing angle, and distance from the operator’s natural hand position—has a direct effect on pick time. A module that is partially obstructed by totes, or mounted at an awkward angle, adds decision time to every pick.
Pushbuttons, Sensors, and Confirmation Devices #
Confirmation devices close the feedback loop. A pushbutton requires a deliberate physical action; a photoelectric sensor or through-beam can confirm a hand pass without extra movement; a barcode scanner adds a verification step. The choice of confirmation device changes the unit time per pick. It also changes the failure mode: worn pushbuttons, misaligned sensors, and dirty optics all create intermittent interruptions that reduce throughput without necessarily producing a fault alarm.
Zone Controllers and Network Topology #
Zone controllers poll or receive events from a bank of light modules and communicate with the host system. The controller’s event-processing rate, the network scan interval, and the host’s transaction latency all contribute to the time between an operator confirmation and the lighting of the next pick. In a correctly functioning system, this delay is perceptible but not limiting. When the host is slow, or the zone controller is overloaded, operators experience “dead time” after each confirmation—a classic bottleneck signature that is often misdiagnosed as an operator issue.
Host Software and Order Release Logic #
The host software determines which orders, lines, and quantities are presented to a zone. Its release logic controls how many active picks exist per zone, whether picks are released as a continuous stream or in discrete waves, and how it handles exceptions such as short picks, substitutions, and location overrides. Capacity planning must include host performance, because the pick-to-light system can only operate as fast as the host feeds it. A fast zone with a slow order-release engine is a system with an artificial ceiling.
Capacity Planning Fundamentals #
Capacity planning for pick-to-light begins with defining the target throughput in a measurable, repeatable way. The two core metrics are order lines per hour and picks per hour. A “pick” is a confirmed interaction at a location; an order line may require multiple picks if the same SKU is stored in multiple locations. It is critical to be explicit about which metric is used in all calculations, because comparing lines per hour against picks per hour creates false conclusions.
Defining Peak Demand and Design Rate #
Start with a demand profile that includes seasonality, daily peaks, and wave structure. The design rate should be set at the sustained peak, not the instantaneous peak. If the operation experiences 30 minutes of extreme activity followed by 90 minutes of low activity, the workstation must sustain the surge without causing upstream or downstream backups. A common planning error is to size a zone for the average hourly rate of a 10-hour shift, which guarantees congestion during the second hour of a peak wave.
Also define the design rate in terms of realized output, not theoretical capacity. Theoretical capacity is the inverse of the shortest per-pick time measured under perfect conditions. Realized output includes rest allowances, exception handling, restocking of consumables, and the unavoidable time spent waiting for the next light. A practical planning factor is to use a sustained pick rate that is 70 to 85 percent of the measured best-observed rate, depending on the complexity of the order mix and the physical layout.
Zone Sizing and Human Factors #
Zone size—the number of active light modules or pick faces assigned to a single operator—must balance walking distance against cognitive load. A very small zone keeps the operator close to all locations but may underutilize the operator’s reach and create long idle periods while waiting for confirmations. A very large zone reduces idle time but increases walking, scanning time, and the chance of misreads. The optimal zone size is an empirical question, but it can be estimated from the average pick-to-pick distance and the confirmation method.
Human factors also determine sustainable capacity. The operator’s reach envelope, the height of dynamic buffers, and the placement of confirmation buttons all influence cycle time. For example, a bottom-level flow rack forces bending; a top-level tote forces stretching or step climbing. Both add time and fatigue. An otherwise well-designed zone with one poorly placed display will disproportionately increase errors and slow the entire line.
Replenishment Interaction #
Pick-to-light zones are typically replenished from the back of a flow rack or by tote exchange. Replenishment is not a maintenance activity; it is a scheduled interruption to picking. Capacity planning must reserve time for replenishment, and bottleneck analysis must distinguish between a picker waiting on a light and a picker waiting because the location was empty. In a correctly planned zone, replenishment is staggered so that no more than a small percentage of pick faces are in a replenishment state at any moment. When replenishment is reactive, it produces a cascade of “no stock” exceptions that look like a hardware problem but are actually an inventory-placement problem.
Conducting a Bottleneck Analysis #
Bottleneck analysis is the process of finding the single constraint that limits zone throughput at a given time. The constraint may shift over the course of a shift, so the analysis must be repeated under different conditions. The goal is not to assign blame but to identify the intervention with the greatest leverage.
Evidence Collection and Observation Methods #
Do not rely on memory or anecdote. Collect time-stamped evidence at a granular level. The most useful data points are:
- Operator confirmation time stamps per zone, aligned to host order-release time stamps.
- Time between consecutive confirmations, measured at the zone controller level.
- Idle gap durations—periods where the operator had no lit module for more than a defined threshold, such as three seconds.
- Replenishment events, including start time, end time, and the SKU and location involved.
- Exception log entries: short picks, location overrides, quantity mismatches, and cancel operations.
- Photographic or video observation of one operator for 15 to 20 minutes, with timestamps, to capture micro-delays that logs cannot show.
When collecting data, align all time sources. Time skew between the host system, the zone controller, and the warehouse management system can make the data unusable. Verify that all timestamps are in the same timezone and that network time synchronization is working.
Common Bottleneck Signatures #
Certain patterns appear repeatedly. The table below links observable symptoms to likely causes and the evidence that differentiates them. Use it as a starting point, not as a final diagnosis.
| Observable Symptom | Likely Constraint | Evidence to Collect | Possible Causes | Next Step |
|---|---|---|---|---|
| Operators wait 3–10 seconds after confirmation before next light | Host order release or zone controller latency | Compare confirmation time to next-light activation time; check host CPU performance at the same minute | Wave release burst, database contention, zone controller polling delay, network congestion | Reproduce with a scripted pick-test during a quiet period; involve controls team |
| One operator is 20–30 percent slower than peers on the same order mix | Individual workstation design or operator technique | Video review; measure pick-to-pick distance; check reach to all modules | Poor module placement, obstructed location, worn pushbutton, non-optimized pick path | Re-position and re-train; check hardware on that specific station |
| Short-pick and “no stock” exceptions increase during peak | Replenishment capacity | Correlate exception time with replenishment schedule; inspect empty locations in log | Insufficient buffer, deferred replenishment, poor forward-pick forecasting | Adjust replenishment trigger points; stagger replenishment tasks |
| All zones slow simultaneously at a regular time of day | Downstream or upstream process | Check conveyor, packing stations, and tote supply for the same time window | Tote jam, packing station full, wave release paused by downstream blockage | Walk the entire flow path; do not modify pick-to-light settings first |
| Errors increase despite no increase in speed | Cognitive load or confirmation feedback | Check whether modules are mislabeled, unreadable, or out of sequence; review error rework costs | Incorrect location labels, ambiguous quantity display, color coding conflict | Audit the physical label-to-module mapping; consult OEM for display configuration |
| Intermittent missing confirmations, but system recovers after a minute | Hardware intermittency, network noise, or power supply | Pull the zone controller log for communication timeouts; inspect connectors and terminations | Loose connector, failing power supply, faulty pushbutton, marginal Ethernet cable | Isolate the zone; perform a controlled continuity check; replace suspected component |
After identifying a likely constraint, verify it with a controlled test. For example, if host release latency is suspected, run a single-zone test with a small batch of orders while monitoring the host’s response time. If the operator’s pick-to-pick time drops significantly when the host is restarted, the evidence supports the hypothesis. A bottleneck analysis is not complete until the proposed fix has been implemented and the measured throughput has been compared against the baseline.
Common Interpretation Errors #
Several recurring mistakes undermine bottleneck analysis in pick-to-light environments. Recognizing them helps prevent wasted effort and wrong interventions.
Confusing pick time with order cycle time. A zone may complete its picks quickly, but the overall order is still late because of waiting time in the queue, packing, or shipping. Pick-to-light throughput is a component of order cycle time, not a substitute for it. Always measure order-level completion time when evaluating the real-world impact of a zone change.
Attributing all idle time to the operator. Idle time is a symptom, not a cause. An operator waiting for a light is responding to a system condition. Before reducing an operator’s assigned zone size, prove that the zone has enough active picks to keep the operator working. Conversely, if an operator habitually starts a pick late after a confirmation, that is an operator-pacing issue, not a hardware issue. The evidence table separates these cases.
Misreading wave release as burst capacity. Releasing an entire wave at once can create a surge of lit modules that overwhelms the zone’s physical capacity, followed by a period of no picks. This is not a failure of the pick-to-light hardware; it is a flaw in release logic. The fix is to throttle release based on zone buffer depth. Controls teams should look for burst patterns before changing hardware settings.
Blindly trusting the average pick rate. Averages hide variance. A zone with a mean pick time of six seconds may alternate between four-second and ten-second picks. If the system is tuned for the mean, the fast picks will not compensate for the slow ones in a way that creates a stable flow. Use the distribution, not the mean, for capacity decisions. At least, use the 85th or 90th percentile of pick time.
Assuming a new system is automatically balanced. Even a newly installed pick-to-light system can be poorly balanced between zones. SKU velocity data, slotting strategy, and order profile all affect zone load. A system that is balanced on day one may be out of balance after a major assortment change or promotional event. Rebalancing is an ongoing operational activity, not a one-time commissioning step.
Maintenance Implications #
Preventive maintenance on pick-to-light systems is often neglected because the failure modes are subtle. A dim display is still readable, a slow pushbutton still works, and an occasional communication timeout disappears on its own. But each of those conditions reduces throughput and increases operator frustration. A structured maintenance program should focus on the interface points between hardware and operator.
Electrical connections are the highest-frequency failure point. Vibration, tote impacts, and frequent reconfiguration of zones loosen connectors and terminal blocks. Loose connections cause intermittent no-light conditions, stray confirms, and communication drops that are notoriously hard to diagnose. Include a scheduled torque check or connection verification on all zone controller terminations and module connectors.
Optical confirmation devices require regular cleaning and alignment. Dust, film residue, and label remnants on a sensor face will eventually produce false confirms or missed confirms. Calibrate or align sensors after any rack movement, module relocation, or impact event. A bent bracket is a common root cause of misalignment that is invisible in a log file.
Physical labels and display overlays must be inspected for fading, peeling, or tearing. A worn label forces the operator to second-guess the location value, which adds a decision pause to every pick. Replace any label that cannot be read clearly from the operator’s normal working position and viewing angle at working distance.
Grounding and power quality matter. Static discharge can reset a module or cause a spurious confirmation. Ensure the racking, workstation frames, and conveyor sections that support pick-to-light modules are properly bonded and grounded according to the site’s electrical standard. Do not run low-voltage signal cables parallel to high-power motor feeders without adequate separation. This is an engineering design issue that should be reviewed during installation and after any electrical modification.
Spare parts strategy should mirror failure probability. Stock replacement pushbuttons, light module faceplates, connectors, and network terminators. For older systems, consider whether the zone controller has a limited service life and plan an upgrade path before a catastrophic failure occurs. In all maintenance work, follow site lockout/tagout procedures and the framework defined in the OEM maintenance manual. Site procedures, lockout requirements, OEM documentation, and competent engineering judgment take priority over the general guidance in this article.
Decision Boundaries and System Evolution #
Capacity planning often reveals that a single zone or controller is undersized. The response may be a tuning change, a hardware addition, or a full system redesign. Knowing where the boundary is between these responses prevents over-engineering and under-engineering.
When to rebalance zones. If total system throughput is acceptable but some zones are saturated while others are underutilized, rebalance by moving SKUs, changing order-to-zone assignment, or adjusting release logic. Rebalancing is an operational decision that typically involves the WMS configuration and slotting team, but not necessarily the OEM. No hardware change is required if the modules and controllers have spare capacity.
When to add a zone or station. If all zones are saturated and the pick-to-light system is confirmed as the system constraint, adding a physical workstation may be justified. Before doing so, verify that the upstream tote presentation and downstream packing processes can handle the additional volume. Adding a pick station without expanding the tote buffer simply moves the bottleneck downstream.
When to involve the OEM. Zone controller firmware behavior, network protocol limits, and the maximum number of modules per controller are defined by the manufacturer. If the diagnosis points to a controller capacity ceiling or a communication protocol limitation, do not attempt to “tune around” it by reducing scan rates or disabling safety features. Contact the OEM for a capacity assessment or upgrade path. Similarly, if the system’s safety logic is involved, do not make changes that could create a hazard. The OEM’s modification procedures and the site’s change-management process take precedence.
When to consider a different technology. Pick-to-light is not the right tool forever. If order profiles become highly variable, if the physical layout changes drastically, or if labor availability makes operator-dependent systems unsustainable, a shift to voice picking, radio-frequency scanning, or a goods-to-person robot may be appropriate. The decision boundary is not technical but economic and operational. The pick-to-light hardware itself is rarely the only factor; the question is whether the entire order-fulfillment architecture can meet the future demand profile.
Safety and authorization boundary. This article is educational and not a substitute for engineering judgment. Any modification that affects safety devices, emergency stops, light curtains, or other protective measures must be performed under the site’s authorized change-control process with OEM approval. Never bypass or defeat a safeguard, and always verify that the system is in a safe state before any servicing. The operations and maintenance teams must clearly document their authority limits so that an operator in the field never has to guess whether a task is within their permission.
Key Takeaways #
- Pick-to-light throughput is determined by the entire chain—host software, release logic, zone controllers, physical layout, confirmation devices, and operator technique—not by the speed of the light modules alone.
- Capacity planning must be based on sustained peak demand and on the distribution of pick times, not the average or the best-observed rate.
- Zone size should be tuned empirically to the operator’s reach envelope, the order profile, and the confirmation method, and it must be re-evaluated after assortment changes.
- Bottleneck analysis requires synchronized time-stamped evidence from the host, the zone controller, and the operator, plus direct observation to capture micro-delays.
- Common misdiagnoses include confusing pick time with order cycle time, treating operator idle time as an operator failure, and ignoring
Related Pearl Gateway Guides #
