Warehouse Operational Excellence, AI & Regional Improvement Portfolio

Business problem

Warehouse orders were processed sequentially even when multiple orders shared the same product, source handling unit and source bin — causing repeated travelling, repeated access to the same locations, more ASRS retrieval cycles and lower picking productivity, with heavy manual effort to identify suitable orders across delivery documents.

Improvement implemented

An AI-assisted Excel/VBA solution that analyses exported SAP warehouse-order data. It:

• Identifies WOs with matching product, source HU and source bin.
• Compares delivery-document numbers and groups WOs across DNs.
• Prevents duplicate WO selection and outputs efficient combinations.
• Allows several DNs to be picked in parallel instead of one at a time.

Business problem

Frequent ASRS interruptions, reset requirements and system faults reduced picking capacity and caused operational delays.

Improvement implemented

• Analysed downtime and reset patterns; improved fault identification and recovery.
• Escalated recurring technical issues using operational data.
• Tracked downtime against reset costs and lost picking productivity.
• Introduced structured monitoring of SRM1, SRM2 and SRM3.

Business problem

When the drop station was not empty, the system condition could leave the load-handling device on hold. Recovery previously took approximately five minutes per occurrence.

Improvement implemented

• Identified the relationship between DS1 occupancy and the ASRS holding condition.
• Supported a resting-data logic correction.
• Reduced recovery time substantially.

Business problem

Inbound employees manually checked handling-unit information, purchase-order details and related records — creating repetitive checking, typing risk, slower verification and inconsistent error detection.

Improvement implemented

The HU Extractor / HU Matcher application:

• HU and PO validation with barcode-based checking.
• Mismatch detection and improved traceability.
• Faster UI and startup in the updated version.

Business problem

VAS records, attachments, billing information and supporting documents involved repetitive manual administration.

Improvement implemented

• Digitised VAS documentation with automated control-number generation.
• Automated folder creation and document storage; attachment handling and PDF generation.
• Improved linkage between VAS work and billing.

Business problem

Temperature data for five warehouse areas — Outbound, Inbound, LPS, MPS and ASRS — was processed manually, one file at a time.

Improvement implemented

A macro that imports files, cleans data, consolidates all five areas, standardises the reporting format and removes manual copying and formatting.

Business problem

Full trolleys, empty trolleys and cage carts lacked clearly designated locations — creating congestion, mixed equipment status, potential obstruction, unclear spider-pickup zones and MHE route concerns.

Improvement implemented

Redesigned Section 1 and Section 2 lane usage, with dedicated locations for empty small/big trolleys, full trolleys, empty and full cage carts, and spider pickup.

Safety consideration: a powered pallet jack crossing a low-traffic walkway was identified as a risk and included in the final layout review.

Business opportunity

Employees and customers need quick answers to repeated questions, but information is spread across websites, documents, work instructions, SOPs and individual subject-matter experts.

Solution developed

An independently designed and developed AI chatbot with:

• Conversational question answering.
• English, Malay, Chinese and Tamil support.
• Website integration and an adaptable knowledge base.

Business opportunity

Commercial teams spend significant time manually searching for prospective companies, checking company information, identifying business gaps, locating contacts and preparing outreach.

Solution developed

An AI Lead Finder application providing:

• Business discovery and lead-list creation.
• Public company-information gathering and opportunity identification.
• Multilingual outreach support, CSV export and structured lead tracking.

Business problem

RFQ and tender preparation involves large amounts of unstructured information, manual document checking, multiple stakeholders, risk of missed requirements, repeated comparison work and difficult handover between commercial and operations.

Solution developed

An RFQ/Tender Assistant that structures tender information, extracts and consolidates key requirements, tracks missing information, organises operational and commercial inputs, and centralises tender information to support response preparation and internal review.