A third-party logistics provider operating a high-volume warehouse in Western Sydney relied on paper pick lists and periodic spreadsheet updates that could not keep pace with same-day dispatch expectations. Small Pea Software delivered a barcode-driven inventory workflow application running on rugged handheld scanners and supervisor web consoles, integrated with their existing WMS for order release and carrier manifest export.
Client Context
The provider stores fast-moving consumer goods for several brand clients with strict cut-off windows for major retailer deliveries. Pick accuracy below client thresholds triggered chargebacks that eroded margins. Floor staff scanned inconsistently — some locations used manufacturer barcodes, others internal location labels handwritten when racks moved. Supervisors discovered pick errors at packing benches, too late for efficient correction.
Leadership had evaluated enterprise WMS modules with substantial licensing and hardware lock-in. They needed targeted workflow software that enforced scan discipline without replacing their WMS investment. Small Pea Software was engaged to build a complementary layer focused on receive, putaway, pick, pack, and cycle count paths with real-time variance alerting.
Problems to Solve
Inventory accuracy audits showed a recurring three percent variance on fast movers, concentrated in partial carton picks and mis-slotted returns. Paper lists did not reflect last-minute order changes from clients, causing pickers to complete obsolete tasks while urgent orders waited reallocation. Cycle counts were ad hoc — entire aisles closed on weekends with overtime costs.
Traceability for batch-sensitive products was inadequate. Clients requesting lot genealogy for quality incidents received reconstructed timelines from paper and camera footage. Training new pickers took weeks because locations were memorised rather than system-guided.
Technical Approach
We built a progressive web application optimised for Android handheld scanners with laser and camera scan inputs, offline queueing for brief connectivity drops within the warehouse, and sync on reconnect. Location barcodes, SKU barcodes, and pick container IDs formed a chained validation — each step required successful scan match before advancing. Supervisor web consoles displayed live pick progress, exception queues, and reassign controls.
Integration used the WMS API for order download and confirmation upload, with a local staging database buffering transactions during outages. Batch and expiry attributes captured at receive propagated through pick confirmation for client traceability exports. Label printing integration generated location and LPN barcodes from templates per client branding rules where required.
Challenges
Warehouse Wi-Fi dead zones near metal racking caused sync delays that initially frustrated pickers. We tuned offline batch sizes, added visual sync indicators, and relocated two access points based on signal heatmaps captured during pilot week. Mixed barcode symbologies required normalisation rules and fuzzy matching with supervisor override logged for audit.
Union of client-specific business rules — some allowed partial picks, others mandated full carton — demanded configurable workflow templates rather than one global path. Peak season volume required load testing at two hundred concurrent scanner sessions; database indexing and optimistic UI updates prevented perceived lag on scan feedback.
Solutions Delivered
Receive workflows enforce scan of inbound ASN lines, capture discrepancies with photo evidence, and suggest putaway locations based on velocity class and client segregation rules. Pick paths optimise walk sequence within aisle constraints exported from WMS wave plans. Pack bench validation rescans contents before carton close, printing SSCC labels integrated to carrier manifest APIs.
Cycle counting shifts to continuous counting by zone with automatic recount triggers when variance exceeds thresholds. Traceability reports assemble lot movement from receive through dispatch with operator and timestamp for client quality teams. Training mode on scanners lets new staff practice without affecting live inventory.
Measurable Outcomes
Pick accuracy improved from ninety-four point two to ninety-nine point one percent within twelve weeks, eliminating most client chargebacks tied to mispicks. Average pick minutes per order line decreased by eighteen percent due to guided paths and elimination of paper reconciliation. Inventory variance on fast movers dropped below zero point eight percent in monthly audits.
Traceability report generation for client inquiries reduced from same-day manual assembly to under ten minutes self-service export. Supervisor overtime on weekend bulk counts fell by roughly forty percent after continuous counting rollout. The provider expanded scanner deployment to a second dock with Small Pea Software supporting template configuration for an additional retail client.
Floor staff feedback shifted from scepticism about "another system" to reliance on scan prompts within three weeks once paper reconciliation disappeared from shift ends. Client quality teams began requesting standardised traceability exports rather than bespoke email threads, which reduced commercial friction during contract renewals. The WMS remained in place — proof that complementary workflow software can deliver accuracy gains without a disruptive platform replacement.