AI & Data Operations: Global Logistics at ABACUS

A consulting presentation for ABACUS Logistics, showcasing data-driven and AI-powered solutions to optimize delivery routes, automate warehouse quality control, and enhance cost efficiency. This deck highlights predictive analytics, computer vision, and real-time optimization to shape the future of global logistics.

Transcript

1. AI & Data Operations CHANGING THE WORLD OF LOGISTICS Presented by Romain JOCHUM, Amy, Tom LERAT, Félix CHAROTTE, GUO Yimeng
2. Table of Contents Points to discuss: ABAKUS Overview Key Operational Challenges Data Landscape & Opportunities AI-Powered Solutions Predicting Delivery Delays Warehousing Quality Control Cost Optimization Implementation Roadmap Expected Benefits
3. ABACUS LOGISTICS Specialises in carriage of goods: domestically and internationally. Most efficient logistics provider leveraging technology & data. Key Activities Vision Key Figures Business Model Step 1: Send Information Step 2: Wait for quote (~8h)
4. Warehouse Quality Control damage, manual inspection Cost Efficiency fuel, labor, static allocation Predicting Delivery Delays traffic, weather, inefficiencies Key Operational Challenges
5. 01 Volume: millions of daily deliveries. 02 Velocity: real-time GPS and weather updates. 03 Variety: structured (logs), semi-structured (ratings), unstructured (camera feeds). 04 Veracity: reliable and accurate data (e.g., GPS, client feedback) 05 Value: insights to improve efficiency (delays, costs, quality) Data Landscape & Opportunities 5Vs Vs
6. Proposed AI Solutions
7. Solutions Description of Response to Tender IDENTIFY TRENDS & ROOT CAUSES ML models to analyze historical data and reveal key delay drivers (traffic peaks, weather patterns, warehouse bottlenecks, regions). AI-POWERED ROUTE OPTIMIZATION Integrate on website for customer to predict time delivery and company to find the fatest way. Optimization algorithm (= Uber) for multi-stop deliveries. Integrates traffic & weather APIs for predictive routing. Suggests best alternative routes in real time. REAL-TIME DASHBOARDS & NOTIFICATIONS Power BI dashboard for managers: live vehicle tracking, delay alerts, performance metrics. Customer notifications: proactive updates if delays are predicted. Predicting Delivery Delays CHALLENGE 1 Traffic, weather, or inefficient route planning. Current systems lack real-time insights and alternative solutions.
8. Predicting Delivery Delays CHALLENGE 1
9. Quality Control CHALLENGE 2 Problems Insights LACKS OF DETAILED DATA There are too little data to accurately infer the root cause of the problem +90,78% DAMAGE BETWEEN WEST - EAST REGION There is a significant range of damage rate between regions - ( 2,48% west / 4,74% east or +90,78%) UNSTANDARDIZED INSPECTION EFFICIENCY (ACROSS WAREHOUSE) Some warehouse successfully manage 94,36% inspection efficiency but some can only manage 70,12%
10. Quality Control CHALLENGE 2 Inspection Efficiency Solution Elimination of Damage Claims 01 Reduced Inspection Time 02 Optimised Truck Loading and Unloading 03
11. Opportunity: ABACUS already has rich datasets (GPS, traffic, customer ratings, historical demand) that can fuel predictive analytics. Solutions : Visualization: Power BI to track satisfaction and delivery performance. Rising costs: Fuel and labor are cutting into margins. Solutions : ML regression, route optimization (OR-Tools, OptaPlanner) Root cause: Resources (trucks, drivers, warehouse staff) are allocated using static schedules instead of real demand patterns. Solutions : Kafka/Flink for real- time traffic & demand streams Cost Optimisation CHALLENGE 3
12. Barriers: system integration, data privacy/security, employee adoption. Pilot projects (single region & warehouse). Implementation Roadmap Phase 1 Consolidate data in cloud + set up AI platform. Phase 2 Phase 3 Global rollout with scaling and continuous improvement. ABACUS_Data.csv POC Algorithms + Dashboard
13. Key Benefits of the solutions Optimize delivery routes for less delays Improve customer satisfaction Reduce operating costs
14. Who you Want to be?
15. Do you have any questions?