SUPPLY CHAIN MANAGEMENT · SESI 05
IT dalam SCM
Information Technology, Business Analytics & AI Workshop
Sistem saraf digital SCM: memahami IT dalam supply chain (ERP, WMS, TMS), business analytics (descriptive, predictive, prescriptive), dan hands-on workshop AI untuk demand forecasting, inventory optimization, dan route optimization.
Tujuan Pembelajaran
IT Fundamentals
Memahami peran strategis IT, enterprise systems (ERP, WMS, TMS), dan emerging technologies.
Memahami peran strategis IT, enterprise systems (ERP, WMS, TMS), dan emerging technologies.
Business Analytics
Membedakan descriptive, predictive, dan prescriptive analytics dalam SCM.
Membedakan descriptive, predictive, dan prescriptive analytics dalam SCM.
AI-Powered Analysis
Menggunakan AI untuk demand forecasting, inventory optimization, dan route optimization.
Menggunakan AI untuk demand forecasting, inventory optimization, dan route optimization.
Data-Driven Culture
Membangun budaya pengambilan keputusan berbasis data.
Membangun budaya pengambilan keputusan berbasis data.
📑 Daftar Isi
1. Peran Strategis IT dalam SCM
IT dalam supply chain adalah infrastruktur digital yang memungkinkan aliran informasi real-time di seluruh rantai pasok – dari supplier sampai customer – untuk mendukung pengambilan keputusan yang cepat, akurat, dan terintegrasi.
4 Peran Strategis IT dalam SCM:
| Peran | Deskripsi | Contoh Aplikasi | Business Impact |
|---|---|---|---|
| Visibility | Memberikan pandangan end-to-end atas seluruh SC | Real-time tracking, control tower, dashboard | Reduce uncertainty, faster response |
| Integration | Menghubungkan sistem dan proses antar fungsi/partner | ERP, EDI, API integration | Eliminate silos, reduce duplication |
| ③ Optimization | Optimasi keputusan operasional dan strategis | Route optimization, inventory optimization | Cost reduction, service improvement |
| ④ Collaboration | Memfasilitasi kolaborasi dengan partner SC | Supplier portal, VMI platform, CPFR | Better coordination, innovation |
🇮🇩 Contoh Indonesia: PT Semen Indonesia
IT transformation journey:
• 2010: Implementasi SAP ERP
• 2015: Tambah WMS untuk warehouse
• 2018: TMS + GPS tracking
• 2020: IoT sensors di pabrik & truk
• 2022: AI-powered demand forecasting
Hasil: Order-to-delivery 7 hari → 3 hari, inventory accuracy 85% → 99%, transportation cost -18%.
IT transformation journey:
• 2010: Implementasi SAP ERP
• 2015: Tambah WMS untuk warehouse
• 2018: TMS + GPS tracking
• 2020: IoT sensors di pabrik & truk
• 2022: AI-powered demand forecasting
Hasil: Order-to-delivery 7 hari → 3 hari, inventory accuracy 85% → 99%, transportation cost -18%.
2. Enterprise Systems: ERP, WMS, TMS, OMS
Enterprise systems adalah tulang punggung digital yang mengintegrasikan berbagai fungsi SCM:
5 Enterprise Systems Utama:
| System | Fungsi Utama | Fitur Kunci | Vendor Populer |
|---|---|---|---|
| ERP (Enterprise Resource Planning) | Integrasi seluruh fungsi bisnis | Single database, real-time reporting, process automation | SAP, Oracle, Microsoft Dynamics |
| WMS (Warehouse Management System) | Optimasi operasi warehouse | Barcode/RFID, bin location, wave picking | Manhattan, Blue Yonder, SAP EWM |
| TMS (Transportation Management) | Optimasi transportasi | Route optimization, carrier selection, tracking | Oracle TMS, SAP TM, BluJay |
| OMS (Order Management System) | Manage order lifecycle | Order orchestration, inventory visibility, returns | IBM Sterling, Manhattan, Oracle |
| APS (Advanced Planning & Scheduling) | Advanced planning | Demand forecasting, production scheduling | Kinaxis, o9 Solutions, Blue Yonder |
🇮🇩 Contoh: PT Unilever Indonesia
System landscape:
• ERP: SAP S/4HANA
• WMS: SAP EWM (15 warehouse)
• TMS: SAP TM + custom route optimization
• APS: Blue Yonder
• Analytics: Power BI + Tableau
Hasil: Order processing 2 jam → 5 menit, forecast accuracy 85%, IT cost -30%.
System landscape:
• ERP: SAP S/4HANA
• WMS: SAP EWM (15 warehouse)
• TMS: SAP TM + custom route optimization
• APS: Blue Yonder
• Analytics: Power BI + Tableau
Hasil: Order processing 2 jam → 5 menit, forecast accuracy 85%, IT cost -30%.
3. Emerging Technologies: IoT, Blockchain, Cloud, Digital Twin
Selain enterprise systems, ada 4 teknologi emerging yang mengubah landscape SCM:
4 Emerging Technologies:
| Teknologi | Definisi | Aplikasi SCM | Contoh Indonesia |
|---|---|---|---|
| 🌐 IoT (Internet of Things) | Jaringan perangkat fisik dengan sensor & konektivitas | Asset tracking, condition monitoring, predictive maintenance | Telkomsel IoT, eFishery IoT, Hexa cold chain |
| ⛓️ Blockchain | Distributed ledger untuk transaksi transparent & immutable | Traceability, smart contracts, provenance verification | IBM Food Trust (Indofood), VeChain |
| ☁️ Cloud Computing | Delivery computing services via internet (SaaS, PaaS, IaaS) | SaaS applications, scalable infrastructure, data lake | AWS, Azure, GCP untuk SCM |
| 🎭 Digital Twin | Virtual replica dari physical assets/processes | Network simulation, what-if analysis, predictive analytics | Siemens, PTC ThingWorx, ANSYS |
🇩 Contoh: PT Pertamina
• 100,000+ IoT sensors di pipa untuk monitoring pressure, temperature
• Digital twin dari pipeline network untuk simulasi flow & leak detection
Hasil: Unplanned downtime -45%, leak detection 24 jam → 5 menit, energy consumption -12%.
• 100,000+ IoT sensors di pipa untuk monitoring pressure, temperature
• Digital twin dari pipeline network untuk simulasi flow & leak detection
Hasil: Unplanned downtime -45%, leak detection 24 jam → 5 menit, energy consumption -12%.
4. Business Analytics: Descriptive, Predictive, Prescriptive
Business Analytics adalah penggunaan data, statistical algorithms, dan machine learning untuk mengidentifikasi insights dan mendukung pengambilan keputusan bisnis.
3 Types of Analytics:
| Type | Pertanyaan | Teknik | Value | Complexity |
|---|---|---|---|---|
| ① Descriptive “What happened?” | Apa yang sudah terjadi? | Reporting, dashboards, data visualization | Low (awareness) | Low |
| ② Predictive “What will happen?” | Apa yang akan terjadi? | Statistical modeling, ML, forecasting | Medium (foresight) | Medium |
| ③ Prescriptive “What should we do?” | Apa yang harus kita lakukan? | Optimization, simulation, decision models | High (action) | High |
🇮 Contoh: PT Astra International
Analytics maturity journey:
• 2010: Level 1-2 (Excel, basic dashboards)
• 2015: Level 2-3 (Power BI, predictive maintenance)
• 2018: Level 3 (ML demand forecasting)
• 2020: Level 3-4 (route optimization)
• 2023: Level 4 (digital twin)
Hasil: Forecast accuracy 65% → 90%, inventory -25%, maintenance cost -35%.
Analytics maturity journey:
• 2010: Level 1-2 (Excel, basic dashboards)
• 2015: Level 2-3 (Power BI, predictive maintenance)
• 2018: Level 3 (ML demand forecasting)
• 2020: Level 3-4 (route optimization)
• 2023: Level 4 (digital twin)
Hasil: Forecast accuracy 65% → 90%, inventory -25%, maintenance cost -35%.
5. Analytics Applications in SCM
10 Key Analytics Applications:
| Application | Analytics Type | Teknik | Business Value |
|---|---|---|---|
| Demand Forecasting | Predictive | Time-series, ML, ensemble | Reduce stockout, optimize inventory |
| Inventory Optimization | Prescriptive | Optimization, simulation | Minimize cost, maximize service |
| Route Optimization | Prescriptive | Vehicle routing, network flow | Reduce cost, improve OTD |
| Supplier Risk Analytics | Predictive | Risk scoring, anomaly detection | Early warning, mitigate disruption |
| Predictive Maintenance | Predictive | IoT data, ML, survival analysis | Reduce downtime & cost |
| Quality Analytics | Predictive | SPC, computer vision | Reduce defects, improve quality |
| Network Design | Prescriptive | Optimization, simulation | Optimal facility location |
| Price Optimization | Prescriptive | Elasticity modeling | Maximize revenue & margin |
| Customer Segmentation | Descriptive/Predictive | Clustering, classification | Tailored service |
| Sustainability Analytics | Descriptive/Predictive | Carbon footprint, LCA | Reduce environmental impact |
🇮🇩 Contoh: PT Indofood CBP
• Demand Forecasting: ML (XGBoost, LSTM) untuk 10,000+ SKU → accuracy 88%
• Inventory Optimization: Multi-echelon → inventory reduction 22%
• Route Optimization: 5,000+ truk → fuel cost -15%
• Predictive Maintenance: IoT sensors → downtime -45%
Total saving: Rp 1.5 triliun/tahun, ROI 400% dalam 3 tahun.
• Demand Forecasting: ML (XGBoost, LSTM) untuk 10,000+ SKU → accuracy 88%
• Inventory Optimization: Multi-echelon → inventory reduction 22%
• Route Optimization: 5,000+ truk → fuel cost -15%
• Predictive Maintenance: IoT sensors → downtime -45%
Total saving: Rp 1.5 triliun/tahun, ROI 400% dalam 3 tahun.
6. Workshop 1: AI-Powered Demand Forecasting
Tugas: Anda Demand Planner di PT Indofood CBP yang harus forecast demand Indomie Goreng untuk 3 bulan ke depan.
📊 Sample Data (12 Bulan):
Month: Jan-Dec 2025
Sales: 125K, 118K, 132K, 128K, 135K, 142K, 148K, 155K, 162K, 158K, 165K, 180K
🎯 Langkah 1: Basic Analysis (5 menit)
📝 Prompt untuk AI:
“Saya punya data penjualan Indomie Goreng 12 bulan. Tolong:
1. Hitung growth rate bulanan dan rata-rata
2. Identifikasi pola musiman
3. Forecast 3 bulan depan dengan 3 metode (naive, moving average, linear trend)
4. Bandingkan akurasi dan rekomendasikan metode terbaik”
“Saya punya data penjualan Indomie Goreng 12 bulan. Tolong:
1. Hitung growth rate bulanan dan rata-rata
2. Identifikasi pola musiman
3. Forecast 3 bulan depan dengan 3 metode (naive, moving average, linear trend)
4. Bandingkan akurasi dan rekomendasikan metode terbaik”
🎯 Langkah 2: Advanced Forecasting (10 menit)
Prompt untuk AI:
“Lakukan advanced forecasting dengan faktor:
– Kompetitor launch produk baru (-5%)
– Promosi Ramadhan (+15%)
– Economic growth 5%
– Pattern: Q4 selalu 20% lebih tinggi dari Q3
Generate Python code dengan Prophet/ARIMA, berikan confidence interval.”
“Lakukan advanced forecasting dengan faktor:
– Kompetitor launch produk baru (-5%)
– Promosi Ramadhan (+15%)
– Economic growth 5%
– Pattern: Q4 selalu 20% lebih tinggi dari Q3
Generate Python code dengan Prophet/ARIMA, berikan confidence interval.”
Contoh Output:
• Jan 2026: 163,400 (-5% competitor)
• Feb 2026: 204,700 (+15% promo)
• Mar 2026: 212,750 (+15% promo)
• Safety Stock: 15% = 25K-32K cartons
• Production Plan: 188K-245K cartons/bulan
• Jan 2026: 163,400 (-5% competitor)
• Feb 2026: 204,700 (+15% promo)
• Mar 2026: 212,750 (+15% promo)
• Safety Stock: 15% = 25K-32K cartons
• Production Plan: 188K-245K cartons/bulan
7. Workshop 2: AI-Powered Inventory Optimization
Tugas: Anda Inventory Manager di PT Unilever Indonesia yang harus optimize inventory untuk 5 SKU utama.
📊 Sample Data:
SKU001: Shampoo 130ml – Demand 500K, Cost Rp 15K, Order Cost Rp 500K, Holding 20%, Lead 14 hari
SKU002: Soap 100g – Demand 800K, Cost Rp 5K, Order Cost Rp 400K, Holding 20%, Lead 10 hari
SKU003: Toothpaste 190g – Demand 600K, Cost Rp 12K, Order Cost Rp 450K, Holding 20%, Lead 12 hari
SKU004: Detergent 900g – Demand 400K, Cost Rp 18K, Order Cost Rp 600K, Holding 20%, Lead 15 hari
SKU005: Deodorant 45ml – Demand 300K, Cost Rp 10K, Order Cost Rp 350K, Holding 20%, Lead 10 hari
Langkah 1: EOQ Calculation (5 menit)
📝 Prompt untuk AI:
“Hitung EOQ untuk setiap SKU: EOQ = sqrt((2 × D × S) / H)
Hitung total annual cost, reorder point, safety stock (service level 95%).
Berikan rekomendasi order frequency.”
“Hitung EOQ untuk setiap SKU: EOQ = sqrt((2 × D × S) / H)
Hitung total annual cost, reorder point, safety stock (service level 95%).
Berikan rekomendasi order frequency.”
🎯 Langkah 2: ABC Analysis (5 menit)
📝 Prompt untuk AI:
“Lakukan ABC analysis berdasarkan annual spend:
– A items (top 70%): tight control, frequent review
– B items (next 20%): moderate control
– C items (remaining 10%): simple control
Berikan inventory policy per kategori.”
“Lakukan ABC analysis berdasarkan annual spend:
– A items (top 70%): tight control, frequent review
– B items (next 20%): moderate control
– C items (remaining 10%): simple control
Berikan inventory policy per kategori.”
Contoh Output:
• ABC Classification:
– A Items (70%): SKU001 (Shampoo), SKU004 (Detergent)
– B Items (20%): SKU003 (Toothpaste)
– C Items (10%): SKU002 (Soap), SKU005 (Deodorant)
• Business Impact: Inventory value -25%, annual cost -18%, ROI 350%.
• ABC Classification:
– A Items (70%): SKU001 (Shampoo), SKU004 (Detergent)
– B Items (20%): SKU003 (Toothpaste)
– C Items (10%): SKU002 (Soap), SKU005 (Deodorant)
• Business Impact: Inventory value -25%, annual cost -18%, ROI 350%.
8. Workshop 3: AI-Powered Route Optimization
Tugas: Anda Logistics Manager di PT Nestle Indonesia yang harus optimize delivery route untuk 1 truk dengan 10 customer di Jakarta.
🎯 Langkah 1: Route Planning (10 menit)
Prompt untuk AI:
“Saya punya 10 customer di Jakarta dengan koordinat GPS, demand, dan time window.
Tolong:
1. Hitung distance matrix (Haversine formula)
2. Solve TSP untuk shortest route
3. Consider time windows & truck capacity (500 cartons)
4. Asumsi speed 30 km/jam, service time 15 menit
5. Berikan optimal sequence, total distance, time, fuel cost”
“Saya punya 10 customer di Jakarta dengan koordinat GPS, demand, dan time window.
Tolong:
1. Hitung distance matrix (Haversine formula)
2. Solve TSP untuk shortest route
3. Consider time windows & truck capacity (500 cartons)
4. Asumsi speed 30 km/jam, service time 15 menit
5. Berikan optimal sequence, total distance, time, fuel cost”
🎯 Langkah 2: Advanced Optimization (10 menit)
📝 Prompt untuk AI:
“Optimize dengan additional constraints:
– Traffic pattern: jam sibuk 07:00-09:00 & 16:00-19:00 (speed -50%)
– Priority customers: D, G, H (harus sebelum 11:00)
– Cluster nearby customers
Berikan final route dengan timeline dan cost saving vs random route.”
“Optimize dengan additional constraints:
– Traffic pattern: jam sibuk 07:00-09:00 & 16:00-19:00 (speed -50%)
– Priority customers: D, G, H (harus sebelum 11:00)
– Cluster nearby customers
Berikan final route dengan timeline dan cost saving vs random route.”
Contoh Output:
• Optimal Route: Warehouse → G → D → I → B → A → H → C → J → F → E
• Total Distance: 71 km (vs 120 km random = 41% shorter)
• Total Time: 6.5 jam (07:00-13:30)
• Fuel Cost: Rp 71,000 (vs Rp 120,000)
• Annual Saving: Rp 12.25 juta per truck (250 days)
• Optimal Route: Warehouse → G → D → I → B → A → H → C → J → F → E
• Total Distance: 71 km (vs 120 km random = 41% shorter)
• Total Time: 6.5 jam (07:00-13:30)
• Fuel Cost: Rp 71,000 (vs Rp 120,000)
• Annual Saving: Rp 12.25 juta per truck (250 days)
Ringkasan Eksekutif Sesi 05
- Peran IT dalam SCM: IT adalah sistem saraf digital – memberikan visibility, integration, optimization, collaboration. Evolusi dari mainframe ke AI-driven.
- Enterprise Systems: 5 sistem utama: ERP, WMS, TMS, OMS, APS. Integration via ESB, iPaaS, API.
- Emerging Technologies: IoT (tracking), Blockchain (traceability), Cloud (scalability), Digital Twin (simulation).
- 3 Types of Analytics: Descriptive (what happened), Predictive (what will happen), Prescriptive (what should we do).
- 10 Analytics Applications: Demand forecasting, inventory optimization, route optimization, supplier risk, predictive maintenance, dll.
- AI-Powered Workshops: Demand forecasting (ML), inventory optimization (EOQ, ABC), route optimization (TSP).
- Data-Driven Culture: 5 pillars: leadership, data literacy, transparency, experimentation, accountability.
- Best Practices: Start with business problem, ensure data quality, validate results, monitor & iterate, ethical use.
Referensi Utama
- Bowersox, D.J., Closs, D.J., & Cooper, M.B. (2019). Supply Chain Logistics Management (5th ed.). McGraw-Hill. Chapter 5.
- Chopra, S., & Meindl, P. (2023). Supply Chain Management: Strategy, Planning, and Operation (8th ed.). Pearson. Chapter 17.
- Davenport, T.H. (2018). The AI Advantage. MIT Press.
- Gartner. (2024). Top Supply Chain Technology Trends.
FAQ IT dalam SCM
Apa perbedaan utama antara ERP, WMS, dan TMS?
ERP mengintegrasikan seluruh fungsi bisnis (finance, HR, procurement, sales). WMS fokus pada optimasi operasi warehouse (receiving, picking, packing). TMS fokus pada optimasi transportasi (route planning, carrier selection, freight audit).
Mengapa data quality penting untuk analytics?
“Garbage in, garbage out.” Sehebat apapun sistem analytics, jika data yang masuk buruk, output yang dihasilkan juga buruk. Cost of poor data quality bisa mencapai 4.5-14% dari revenue.
Apa perbedaan descriptive, predictive, dan prescriptive analytics?
Descriptive menjawab “what happened” (reporting, dashboards). Predictive menjawab “what will happen” (forecasting, ML). Prescriptive menjawab “what should we do” (optimization, simulation).
Bagaimana cara memulai AI-powered analytics di SCM?
Mulai dengan business problem yang jelas, pastikan data quality baik, mulai dari simple models (Excel, basic ML), validate results dengan domain knowledge, lalu scale gradually ke advanced models.