Order picking is the process of retrieving specific items from warehouse storage to fulfil customer orders. It is consistently identified as the most labour-intensive, error-prone, and costly activity in warehouse operations — accounting for 40–65% of total warehouse operating cost in most facilities.

Automated order picking systems reduce this cost by directing operators with precision (pick-to-light, voice picking), bringing inventory to operators automatically (goods-to-person), or eliminating operators from the picking zone entirely (robotic picking). In Malaysia’s manufacturing and distribution sectors, where order complexity is rising and skilled warehouse labour is increasingly scarce, picking automation has become a high-priority investment.

DNC Automation designs and integrates automated order picking systems that match the specific inventory profile, order pattern, and throughput requirements of Malaysian customers.

The Cost of Manual Picking

Before examining automated alternatives, the cost structure of manual picking must be understood:

Walking time: In a conventional pick-from-shelving operation, a picker walks 15–25 kilometres per shift searching for and travelling to pick locations. Walking consumes 40–60% of total picker time.

Search and confirm time: Identifying the correct product at each pick location — reading labels, checking quantities — adds a further 15–20% of picker time.

Error rates: Manual picking without systematic guidance achieves order accuracy rates of 97–99.5%, meaning 5–30 errors per 1,000 pick lines. Each picking error costs RM 50–500 in returns processing, re-picking, and customer service.

Labour dependency: Manual picking operations are directly dependent on headcount. Throughput cannot increase without hiring more pickers — a challenge in Malaysia’s tight industrial labour market.

Automated picking systems address all four of these cost drivers simultaneously.

Types of Automated Order Picking Systems

Pick-to-Light Systems

Pick-to-light uses LED light indicators mounted on each storage location. When an order is processed, lights illuminate at the locations where picks are required, showing the operator exactly where to pick and how many units to take. The operator picks the indicated quantity and presses a confirmation button on the light module.

Key characteristics:

• Operator walks to pick locations (man-to-goods), but search time is eliminated
• Pick accuracy improves to 99.9–99.99% (lights eliminate mis-identification)
• Throughput increases 30–50% compared to paper-based picking
• Works in ambient and chilled environments
• Best suited to high-velocity, limited-SKU zones (e.g., fast-moving consumer goods, automotive replenishment)

Malaysian applications: Pick-to-light is widely used in Selangor automotive parts distribution centres and F&B mixed-case picking operations.

Voice Picking Systems

Voice picking uses a headset with a speech recognition system. The WMS sends pick instructions to the operator’s headset; the operator hears “Go to location A-03-07, pick 4 cases of product 1234, confirm.” The operator confirms by speaking a check digit, and the WMS records the confirmed pick.

Key characteristics:

• Hands-free and eyes-free operation (operator’s hands are free to pick, eyes are on inventory)
• Works in cold, dark, or gloved environments where screen-based systems are difficult
• Language flexibility — systems can operate in Bahasa Malaysia, English, or Mandarin
• Accuracy rates of 99.9%+
• Best suited to pallet and case picking in ambient or cold storage

Malaysian applications: Voice picking is the standard picking technology in large Malaysian cold storage distribution centres (frozen food, pharmaceutical cold chain) where gloves and cold environment prevent touch-screen operation.

Put-to-Light (Sortation to Order)

Put-to-light is the reverse of pick-to-light — used for order assembly rather than picking. A single item is picked in bulk and then sorted into individual customer orders. Lights on order cartons or lanes indicate which order each item belongs to.

Application: Pharmaceutical distributors and consumer goods companies with many small orders (clinic replenishment, convenience store delivery) use put-to-light to assemble mixed-product orders from bulk picks.

Goods-to-Person (GTP) Automated Picking

In goods-to-person systems, the inventory comes to the operator rather than the operator travelling to the inventory. ASRS systems (VLMs, carousels, mini-load cranes, shuttle systems) deliver storage totes or trays to fixed ergonomic picking workstations.

Key characteristics:

• Walking time eliminated entirely — operators stand at fixed workstations
• Highest picking throughput per operator (300–600 picks per hour in well-designed GTP systems)
• Can be combined with pick-to-light at the workstation for additional accuracy guidance
• Highest capital investment of any picking technology
• Best suited to high-SKU, high-order-volume operations (e-commerce fulfilment, pharmaceutical distribution)

Malaysian applications: Electronics distributors in Penang use GTP with mini-load ASRS for high-volume small-part picking. DNC Automation has designed GTP picking workstations integrated with VLM and carousel systems.

These technologies can be broadly categorized into 4 primary types

Robotic Picking

Robotic picking systems use articulated robot arms or collaborative robots (cobots) equipped with vision systems and flexible end-of-arm tooling to pick individual items from storage and place them into order containers.

Key characteristics:

• Fully unattended operation — no human picker in the pick zone
• Suitable for structured pick environments with defined item shapes (cartons, bottles, cans)
• 24/7 operation without fatigue, breaks, or turnover
• Current limitations: complex shapes, soft/flexible items, and highly mixed SKU profiles require advanced vision systems
• Investment: RM 500,000–2M+ per picking cell depending on complexity

Malaysian applications: Robotic picking is at an early adoption stage in Malaysia, primarily in pharmaceutical and consumer goods companies as pilot projects. Full deployment is growing as system costs decrease.

Autonomous Mobile Robot (AMR) Picking

AMR picking systems deploy a fleet of mobile robots that travel through warehouse aisles. Operators walk short distances within their assigned zone while the AMR robot follows them, carrying the order tote. The WMS directs the operator to the next pick; the robot follows.

Key characteristics:

• Lower capital cost than full GTP systems (no fixed ASRS infrastructure)
• Scalable — add more robots to increase throughput
• Flexible — system can operate in existing warehouse layout without racking changes
• Reduces operator walking by 60–80%

Malaysian applications: AMR picking is being adopted by large F&B distributors and 3PL operators in Malaysia as a first step into warehouse automation.

Picking System Selection Framework

Matching the right picking technology to an operation requires analysis of four parameters:

1. Order Profile

• Few SKUs, high volume per pick: Pick-to-light or voice picking in static storage
• Many SKUs, low volume per pick: Goods-to-person (VLM, carousel, or mini-load)
• Mixed case orders with many order lines: Voice picking or AMR-assisted picking
• Small, many-SKU individual orders: Goods-to-person robotic picking

2. Throughput Requirement

3. Environment

Cold storage, wet, or gloved environments favour voice picking over touch-screen or light-based systems. High-dust factory environments require ruggedised equipment. Pharmaceutical clean rooms require specific equipment certifications.

4. Existing Infrastructure

Operations with existing ASRS or conveyor systems can extend them to support goods-to-person picking at lower incremental cost. Greenfield operations have full flexibility to design the optimal picking architecture.

WMS Integration for Picking Systems

All picking automation systems must interface with a warehouse management system that:

• Releases pick orders and assigns them to picking zones or operators
• Sequences picks to minimise travel (wave planning, zone routing)
• Confirms completed picks and updates inventory in real time
• Manages replenishment of pick faces from bulk storage
• Tracks operator productivity (picks per hour, error rate, shift throughput)

DNC Automation’s software team integrates picking systems with leading WMS and ERP platforms including SAP, Oracle, and Microsoft Dynamics 365.

5 key benefits that make automated picking technology a critical investment

Picking System ROI in Malaysia

Voice Picking ROI (Case Study — Malaysian Cold Store)

A Malaysian frozen food distributor with 80 pickers across three shifts implemented voice picking:

• Picking productivity improved from 90 picks/hour to 155 picks/hour (+72%)
• Picking accuracy improved from 98.8% to 99.92%
• Headcount reduced from 80 to 52 pickers (28 redeployments/redundancies over 18 months)
• Annual labour saving: RM 840,000
• Voice system investment: RM 380,000
• Payback: 5.4 months

Pick-to-Light ROI (Automotive Parts Distribution)

An automotive parts distributor in Shah Alam implemented pick-to-light in their fast-moving pick zone (top 500 SKUs by velocity):

• Error rate reduced from 1.2% to 0.05%
• Return processing cost reduced by RM 280,000/year
• Productivity increased 40% in the lit zone
• Pick-to-light investment: RM 220,000
• Payback: 9.4 months

Goods-to-Person ROI (Electronics Distributor, Penang)

A Penang electronics distributor installed a 4-VLM pod with conveyor integration:

• Floor space for small-parts storage reduced by 65%
• Picking accuracy improved from 98.5% to 99.97%
• Picker headcount reduced from 12 to 6 in the automated zone
• Annual saving (labour + space): RM 560,000
• Investment: RM 1.4M
• Payback: 2.5 years

DNC Automation’s Picking System Services

DNC Automation designs integrated picking automation solutions for Malaysian manufacturers and distributors:

Picking system design: Analysis of order profile, SKU velocity, throughput requirements, and environment to recommend the optimal picking technology mix.

Hardware supply and installation: Pick-to-light hardware, voice picking headsets and servers, conveyor workstations, AMR robots, and goods-to-person storage systems.

WMS and ERP integration: Connection to existing inventory management platforms or supply of standalone WMS with picking modules.

Operator training: System operation training, accuracy target-setting, and productivity KPI monitoring.

Performance monitoring: Post-go-live productivity dashboards and accuracy reporting, with DNC support team review of underperforming metrics.

Frequently Asked Questions

Which picking system achieves the highest accuracy?

Goods-to-person systems (VLM, mini-load, carousel) combined with pick-to-light confirmation at the workstation achieve the highest accuracy — typically 99.97–99.99%. Voice picking achieves 99.9%+ in cold and case-picking environments. All are significantly better than paper-based picking.

Can picking automation handle fragile products?

Voice picking, pick-to-light, and AMR-assisted systems handle any product that operators can pick manually. Robotic picking of fragile or irregular items requires advanced end-of-arm tooling and vision systems — feasible but at higher cost.

How do picking systems handle peak demand?

Voice picking and pick-to-light systems scale by adding operators and headsets. GTP systems scale by increasing wave frequency and operating hours. AMR fleets scale by deploying additional robots. Most systems can handle 2–3× normal throughput during peak periods with operational adjustments.

What is the minimum order volume to justify picking automation?

Voice picking and pick-to-light have low per-operator capital cost (RM 5,000–15,000 per operator) and justify investment in operations picking 500+ order lines per day. GTP and robotic systems require higher volumes to justify their capital cost.

Conclusion

Automated order picking systems — voice picking, pick-to-light, goods-to-person, and robotic picking — deliver measurable improvements in picking accuracy, throughput, and labour efficiency for Malaysian warehouse operations. The right system depends on order profile, throughput requirement, environment, and investment budget.

DNC Automation Malaysia designs picking automation solutions that integrate with existing warehouse infrastructure and business systems, delivering sustainable improvements in operational performance.

Contact DNC Automation Malaysia to discuss your picking automation requirements and receive a system recommendation for your operation.

Every picking operation updates inventory instantly across all connected systems

Picking System Implementation: From Procurement to Go-Live

Phase 1: Picking System Assessment (2–4 weeks)

Before selecting a picking technology, DNC Automation conducts an operational assessment covering:

• Current picking productivity (picks per operator per hour by product zone)
• Current error rate by zone and product type
• Order line volume and composition (single-pick orders vs. multi-line)
• Peak-to-average throughput ratio
• Environmental constraints (temperature, PPE requirements, noise level)
• Existing WMS capabilities and integration readiness

This assessment produces a technology recommendation with ROI model before any equipment commitment.

Phase 2: System Design (4–8 weeks)

Picking system design covers:

• Zone definition — which products are handled by which technology
• Hardware layout — pick face design, light module placement, workstation ergonomics
• WMS configuration — wave planning rules, pick sequence logic, confirmation requirements
• Integration design — WMS-to-picking-system interface specification
• Staff change management plan — how current pickers are transitioned to new system

Phase 3: Installation and Configuration (3–8 weeks)

Hardware installation (pick-to-light modules, voice system servers, AMR fleet deployment), followed by WMS configuration and end-to-end system testing. DNC Automation conducts pre-go-live performance verification against defined throughput and accuracy targets.

Phase 4: Operator Training (1–2 weeks)

Effective picking system training accelerates productivity ramp-up and reduces the post-go-live performance dip that poorly trained teams experience:

• Voice picking: operators trained on command phrases, check digit confirmation, exception handling (can’t find item, wrong quantity in location)
• Pick-to-light: operators trained on light interpretation, quantity adjustment, and multi-order picking sequences
• GTP workstation: operators trained on tray/tote handling, pick confirmation, and replenishment call procedures

Phase 5: Go-Live and Ramp-Up (4–8 weeks)

A controlled go-live sequence: start with one zone or shift, verify performance, expand to full operation. DNC Automation provides on-site support during the first 2–4 weeks post-go-live to resolve configuration issues and support operators through the learning curve.

Picking System Performance Management

Picking automation delivers its full value only when performance is actively managed. DNC Automation provides post-go-live performance management tools:

KPI Dashboard

Real-time dashboard showing:

• Current picks per operator per hour vs. target
• Error rate by operator, zone, and product
• System utilisation rate (for GTP systems — % of time goods are presenting at workstations)
• Wave completion rate vs. shipping cutoff

Operator Performance Coaching

Voice picking and pick-to-light systems capture individual operator performance data. Supervisors can identify underperforming operators early and provide targeted coaching — a capability impossible in paper-based operations.

Wave Optimisation

Over time, DNC Automation analyses wave planning data to identify opportunities for improved wave composition (fewer operator changeovers between zones), better pick sequence logic (reduced travel in man-to-goods zones), and throughput balancing across workstations.

Malaysian Industry Standards for Picking Accuracy

Different industries in Malaysia have different minimum picking accuracy requirements driven by regulatory, contractual, or commercial standards:

Manual picking typically achieves 97–99.5% in well-managed operations — below the minimum for pharmaceutical, automotive, and premium retail supply chains. Automated picking systems meet all of the above standards when properly implemented.

Picking Automation and Malaysian Labour Law

Implementing picking automation that reduces headcount must be managed in compliance with Malaysia’s Employment Act 1955 and Industrial Relations Act 1967:

• Redundancy resulting from automation is a recognised basis for retrenchment under Malaysian law, but requires proper notice periods, severance pay calculation, and in some cases, notification to the Department of Labour
• Best practice: retrain displaced pickers for system operator or maintenance roles where possible; phase implementation to allow natural attrition to absorb headcount reduction before formal retrenchment

DNC Automation does not provide employment law advice but recommends customers engage their HR consultants and the Malaysian Institute of Human Resource Management (MIHRM) when planning workforce transitions alongside automation implementation.

Systems cost RM 1-8 million depending on scale and technology.

Conclusion (Extended)

Automated order picking systems represent the most impactful single investment most Malaysian warehouse operations can make in productivity and accuracy. The technology is proven, the ROI is rapid (often under 2 years for voice picking and pick-to-light), and the operational risk of implementation is manageable with a structured approach.

DNC Automation Malaysia designs picking automation solutions that integrate with your existing WMS, conveyor, and ASRS infrastructure — delivering sustainable picking performance improvement from day one of live operation.

Contact DNC Automation Malaysia to discuss picking automation for your warehouse and receive a throughput and accuracy improvement projection for your specific operation.

Order Picking Technology Selection: Decision Framework

Selecting the right order picking technology requires matching the technology’s capabilities to your operational profile. No single technology is optimal for all operations.

The SKU Velocity Matrix

Map your SKU portfolio against two dimensions: pick frequency (picks per day) and SKU count:

High frequency, low SKU count (ASRS goods-to-person or pick-to-light):

Operations with 50–500 SKUs each picked many times per day benefit from goods-to-person systems where ASRS delivers the SKU to the operator. The operator never walks — eliminating 40–60% of pick cycle time. Malaysian FMCG distributors typically fall in this category.

High frequency, high SKU count (voice or goods-to-person with large storage):

Operations with 5,000–50,000 SKUs and high overall pick velocity require a combination: fast-moving SKUs in goods-to-person ASRS, slow-moving SKUs in pick-by-voice or pick-to-light zones. Pharmaceutical distributors with 30,000+ drug codes use this hybrid approach.

Low frequency, high SKU count (voice or conventional picking):

Low-velocity operations with many SKUs may not justify the capital investment of goods-to-person ASRS. Voice-directed picking with a conventional WMS delivers significant accuracy improvement at lower cost.

Low frequency, low SKU count (minimal technology needed):

Small operations with few SKUs and low throughput may achieve adequate results with standard barcode scan confirmation without additional automation.

Integration with ASRS

When order picking connects to ASRS, the system design must address:

Goods-to-person sequencing: The WCS sequences ASRS deliveries to picking workstations in the order that minimises crane/shuttle travel — not the order that the operator would naturally pick. This WCS optimisation improves throughput by 20–35% compared to sequential pick-by-order approaches.

Batch picking support: For operations processing many small orders (e-commerce), batch picking collects items for multiple orders simultaneously and then sorts them by order at a downstream put-to-light station. The ASRS must support batch wave release — delivering totes containing items for multiple orders rather than one order at a time.

Exception handling: When the ASRS delivers a tote with a quantity discrepancy (damaged item, short pick by previous operator), the picking workstation must have a process to report the exception and trigger ASRS inventory correction — without stopping the pick wave.

DNC Automation designs the complete picking workflow — from WMS wave release through ASRS sequencing to pick confirmation — as an integrated system, not separate vendor components.

Pick Accuracy: Measurement and Continuous Improvement

Defining picking accuracy: Pick accuracy is typically measured as percentage of order lines picked correctly (correct item, correct quantity, to correct destination). Industry standard measurement:

• 1,000 order lines audited per month by physical verification
• Error rate = number of incorrect lines / total lines picked × 100%

Baseline by technology:

• Manual pick-and-check (no technology): 96–98%
• Voice picking: 99.5–99.8%
• Pick-to-light: 99.7–99.9%
• Goods-to-person robotic: 99.95–99.99%

Cost of a picking error: In Malaysian B2B operations, a picking error typically costs RM 150–500 per incident (including: credit note processing, return logistics, replacement shipment, customer service time, and potential customer relationship damage). At 98% accuracy with 1,000 picks per day, 20 errors per day × RM 200 average cost = RM 4,000/day = RM 1,460,000/year in error costs.

Improving to 99.8% accuracy reduces errors from 20/day to 2/day — a saving of RM 1,314,000/year. This accuracy improvement alone often funds the picking automation investment.

Continuous Improvement Process

After automated picking is live, DNC Automation recommends a structured accuracy improvement process:

1. Weekly error log review: Categorise all picking errors by type (wrong item, wrong quantity, wrong destination) and identify systemic patterns
2. Root cause analysis: For each error category, identify whether the error is a WMS configuration issue (wrong pick quantity), a labelling issue (barcode not readable), a training issue (operator procedure not followed), or a system issue (pick-to-light display unclear)
3. Corrective action: Implement specific fixes for each root cause — this is more effective than general retraining
4. Monthly accuracy trend report: Track accuracy improvement over time and report to management as a KPI

Most operations achieve best-practice accuracy (99.9%+) within 60–90 days of live operation as systemic issues are identified and resolved.

Regulatory Compliance in Automated Order Picking

Malaysian manufacturers in regulated industries must ensure that automated picking systems support compliance requirements — not just operational efficiency.

Pharmaceutical serialisation (NPRA): For licensed pharmaceutical distributors, each picked unit must have its serialisation code confirmed at pick. The picking system scans or verifies the 2D barcode on the pharmaceutical package, confirms the serial number matches the WMS pick instruction, and records the confirmation for NPRA audit. Robotic picking with vision verification does this automatically.

Halal confirmation (JAKIM): In automated facilities with halal and non-halal products, the WMS must enforce that goods-to-person ASRS delivers only halal products to halal picking stations. Pick-to-light systems must display alerts if a non-halal product is accidentally presented to a halal pick zone.

Food traceability (JKDM): Automated picking with WMS lot tracking provides complete outbound traceability — which specific lot was picked for which customer order, at what time. This data is exportable for JKDM food safety audit and supports recall management.

DNC Automation configures automated picking systems with industry-specific compliance controls as part of the standard implementation package for regulated industries.