shelf-life-management

保质期管理

你是保质期管理和易腐产品供应链优化方面的专家。你的目标是通过管理到期日期，帮助减少浪费、最大限度保持新鲜度、优化库存周转并确保产品质量。

初步评估

实施保质期管理前，先了解以下信息：

1. 产品特性

   • 哪些产品需要关注保质期？（食品、药品、化妆品）
   • 保质期有多长？（天、周、月）
   • 存储要求是什么？（常温、冷藏、冷冻）
   • 有哪些监管要求？（FDA、USDA、欧盟法规）
   • 日期代码格式是什么？（食用期限、销售期限、最佳食用期、生产日期）

2. 当前状况

   • 当前的浪费／变质率是多少？（占库存的百分比）
   • 采用哪种库存周转方法？（FIFO、FEFO、人工）
   • 是否具备日期代码追踪能力？（WMS、人工）
   • 是否有降价／清仓流程？
   • 客户是否投诉产品不够新鲜？

3. 供应链特征

   • 从生产到上架的交货周期有多长？
   • 有多少个节点（工厂、DC、门店）？
   • 补货频率如何？
   • 促销活动会产生什么影响？

4. 业务影响

   • 年度浪费成本是多少（变质＋降价）？
   • 缺货造成了多少销售损失？
   • 是否存在客户满意度问题？
   • 是否有合规处罚或召回？

---

保质期管理框架

保质期定义

关键日期类型：

1. 生产日期

   • 产品生产的日期
   • 保质期计算的起点

2. 到期日／食用期限

   • 应使用／食用产品的最后日期
   • 涉及安全问题（尤其是食品、药品）
   • 法规要求

3. 最佳食用期

   • 品质日期（不代表安全期限）
   • 产品可能仍然安全，但品质会下降
   • 食品中常见

4. 销售期限

   • 零售商应销售产品的最后日期
   • 在到期日前预留缓冲时间
   • 典型计算方式：到期日减去 X 天

剩余保质期（RSL）：

RSL = Expiration Date - Current Date
RSL % = (Expiration Date - Current Date) / (Expiration Date - Manufacturing Date) × 100

保质期区间

import pandas as pd
import numpy as np
from datetime import datetime, timedelta

class ShelfLifeManager:
    """
    Manage shelf life and expiration dates
    """

    def __init__(self, shelf_life_days):
        self.shelf_life_days = shelf_life_days

        # Define shelf life zones
        self.zones = {
            'green': {'min_pct': 67, 'max_pct': 100, 'action': 'Normal sales'},
            'yellow': {'min_pct': 33, 'max_pct': 67, 'action': 'Priority sales'},
            'red': {'min_pct': 10, 'max_pct': 33, 'action': 'Markdown/clearance'},
            'expired': {'min_pct': 0, 'max_pct': 10, 'action': 'Pull from shelf'}
        }

    def calculate_rsl(self, manufacturing_date, current_date=None):
        """Calculate remaining shelf life"""

        if current_date is None:
            current_date = datetime.now()

        # Convert to datetime if strings
        if isinstance(manufacturing_date, str):
            manufacturing_date = pd.to_datetime(manufacturing_date)
        if isinstance(current_date, str):
            current_date = pd.to_datetime(current_date)

        expiration_date = manufacturing_date + timedelta(days=self.shelf_life_days)
        rsl_days = (expiration_date - current_date).days
        rsl_pct = (rsl_days / self.shelf_life_days) * 100

        return {
            'manufacturing_date': manufacturing_date,
            'expiration_date': expiration_date,
            'current_date': current_date,
            'rsl_days': max(0, rsl_days),
            'rsl_pct': max(0, rsl_pct),
            'expired': rsl_days <= 0
        }

    def classify_zone(self, rsl_pct):
        """Classify product into shelf life zone"""

        for zone_name, zone_info in self.zones.items():
            if zone_info['min_pct'] <= rsl_pct < zone_info['max_pct']:
                return {
                    'zone': zone_name,
                    'action': zone_info['action']
                }

        return {'zone': 'expired', 'action': 'Pull from shelf'}

    def generate_shelf_life_report(self, inventory_df):
        """
        Generate shelf life report for inventory

        Parameters:
        - inventory_df: DataFrame with columns ['sku', 'lot', 'manufacturing_date',
                       'quantity', 'location']

        Returns:
        - report with expiration analysis
        """

        current_date = datetime.now()

        # Calculate RSL for each lot
        inventory_df['rsl_info'] = inventory_df['manufacturing_date'].apply(
            lambda x: self.calculate_rsl(x, current_date)
        )

        # Extract RSL values
        inventory_df['rsl_days'] = inventory_df['rsl_info'].apply(lambda x: x['rsl_days'])
        inventory_df['rsl_pct'] = inventory_df['rsl_info'].apply(lambda x: x['rsl_pct'])
        inventory_df['expiration_date'] = inventory_df['rsl_info'].apply(
            lambda x: x['expiration_date']
        )
        inventory_df['expired'] = inventory_df['rsl_info'].apply(lambda x: x['expired'])

        # Classify zones
        inventory_df['zone_info'] = inventory_df['rsl_pct'].apply(self.classify_zone)
        inventory_df['zone'] = inventory_df['zone_info'].apply(lambda x: x['zone'])
        inventory_df['action'] = inventory_df['zone_info'].apply(lambda x: x['action'])

        # Summary by zone
        zone_summary = inventory_df.groupby('zone').agg({
            'quantity': 'sum',
            'lot': 'count'
        }).rename(columns={'lot': 'num_lots'})

        # Expiring soon (next 7 days)
        expiring_soon = inventory_df[
            (inventory_df['rsl_days'] <= 7) &
            (inventory_df['rsl_days'] > 0)
        ]

        # Expired inventory
        expired_inventory = inventory_df[inventory_df['expired'] == True]

        report = {
            'total_inventory': inventory_df['quantity'].sum(),
            'total_lots': len(inventory_df),
            'zone_summary': zone_summary,
            'expiring_soon_7days': {
                'quantity': expiring_soon['quantity'].sum(),
                'lots': len(expiring_soon),
                'details': expiring_soon[['sku', 'lot', 'quantity', 'rsl_days', 'location']]
            },
            'expired': {
                'quantity': expired_inventory['quantity'].sum(),
                'lots': len(expired_inventory),
                'details': expired_inventory[['sku', 'lot', 'quantity', 'expiration_date', 'location']]
            }
        }

        return report


# Example usage
manager = ShelfLifeManager(shelf_life_days=120)  # 120-day shelf life

inventory = pd.DataFrame({
    'sku': ['SKU_A', 'SKU_A', 'SKU_A', 'SKU_B', 'SKU_B'],
    'lot': ['LOT001', 'LOT002', 'LOT003', 'LOT004', 'LOT005'],
    'manufacturing_date': [
        datetime.now() - timedelta(days=100),  # Old
        datetime.now() - timedelta(days=60),   # Medium
        datetime.now() - timedelta(days=10),   # Fresh
        datetime.now() - timedelta(days=125),  # Expired
        datetime.now() - timedelta(days=80)    # Medium
    ],
    'quantity': [500, 1000, 1500, 200, 800],
    'location': ['DC1', 'DC1', 'DC2', 'DC1', 'DC2']
})

report = manager.generate_shelf_life_report(inventory)

print("Zone Summary:")
print(report['zone_summary'])
print(f"\nExpiring in 7 days: {report['expiring_soon_7days']['quantity']} units")
print(f"Expired: {report['expired']['quantity']} units")

---

FEFO（先到期先出）实施

FEFO 分配逻辑

class FEFOInventoryManager:
    """
    Implement FEFO (First-Expired-First-Out) inventory allocation
    """

    def __init__(self, inventory_df):
        """
        Initialize with inventory

        Parameters:
        - inventory_df: DataFrame with columns ['sku', 'lot', 'expiration_date',
                       'quantity', 'location']
        """
        self.inventory = inventory_df.copy()

    def allocate_order(self, sku, quantity_needed, location=None,
                        min_rsl_days=None):
        """
        Allocate inventory using FEFO logic

        Parameters:
        - sku: product SKU
        - quantity_needed: quantity to allocate
        - location: preferred location (None = any)
        - min_rsl_days: minimum remaining shelf life (customer requirement)

        Returns:
        - allocation list of lots
        """

        # Filter to SKU
        available = self.inventory[
            (self.inventory['sku'] == sku) &
            (self.inventory['quantity'] > 0)
        ].copy()

        # Filter by location if specified
        if location:
            available = available[available['location'] == location]

        # Filter by minimum RSL if specified
        if min_rsl_days:
            current_date = datetime.now()
            available = available[
                (available['expiration_date'] - current_date).dt.days >= min_rsl_days
            ]

        # Sort by expiration date (earliest first) - FEFO
        available = available.sort_values('expiration_date')

        # Allocate
        allocation = []
        remaining_need = quantity_needed

        for idx, row in available.iterrows():
            if remaining_need <= 0:
                break

            # Allocate from this lot
            allocate_qty = min(remaining_need, row['quantity'])

            allocation.append({
                'sku': sku,
                'lot': row['lot'],
                'location': row['location'],
                'expiration_date': row['expiration_date'],
                'quantity': allocate_qty,
                'rsl_days': (row['expiration_date'] - datetime.now()).days
            })

            # Update remaining need
            remaining_need -= allocate_qty

            # Update inventory
            self.inventory.loc[idx, 'quantity'] -= allocate_qty

        # Check if fully allocated
        allocated_qty = sum(a['quantity'] for a in allocation)
        shortage = quantity_needed - allocated_qty

        return {
            'allocated': allocation,
            'total_allocated': allocated_qty,
            'shortage': shortage,
            'fill_rate': allocated_qty / quantity_needed if quantity_needed > 0 else 0
        }

    def get_inventory_summary(self):
        """Get current inventory summary"""

        summary = self.inventory.groupby(['sku', 'location']).agg({
            'quantity': 'sum',
            'lot': 'count',
            'expiration_date': ['min', 'max']
        })

        return summary


# Example
inventory = pd.DataFrame({
    'sku': ['SKU_A', 'SKU_A', 'SKU_A', 'SKU_A'],
    'lot': ['LOT001', 'LOT002', 'LOT003', 'LOT004'],
    'expiration_date': pd.to_datetime([
        '2025-03-15',
        '2025-04-20',
        '2025-02-10',  # Oldest - should allocate first
        '2025-05-01'
    ]),
    'quantity': [500, 800, 300, 1000],
    'location': ['DC1', 'DC1', 'DC1', 'DC2']
})

fefo = FEFOInventoryManager(inventory)

# Allocate order
order = fefo.allocate_order(
    sku='SKU_A',
    quantity_needed=1000,
    location='DC1',
    min_rsl_days=30  # Customer requires 30 days min shelf life
)

print("Allocation:")
for alloc in order['allocated']:
    print(f"  Lot {alloc['lot']}: {alloc['quantity']} units, "
          f"RSL: {alloc['rsl_days']} days")

print(f"\nTotal Allocated: {order['total_allocated']}")
print(f"Shortage: {order['shortage']}")

---

减少浪费的策略

动态降价优化

import numpy as np
from scipy.optimize import minimize_scalar

def optimize_markdown_timing(current_rsl_days, regular_price, cost,
                              demand_elasticity=-2.0):
    """
    Optimize when to markdown product to minimize waste

    Parameters:
    - current_rsl_days: remaining shelf life
    - regular_price: normal selling price
    - cost: product cost
    - demand_elasticity: price elasticity of demand

    Returns:
    - optimal markdown timing and price
    """

    def expected_profit(markdown_day):
        """Calculate expected profit if markdown starts on given day"""

        # Days at full price
        days_full_price = min(markdown_day, current_rsl_days)

        # Days at markdown price
        days_markdown = max(0, current_rsl_days - markdown_day)

        # Demand curves (simplified)
        daily_demand_full = 10  # Base demand at full price
        markdown_pct = min(0.5, days_markdown / current_rsl_days)  # Up to 50% off
        markdown_price = regular_price * (1 - markdown_pct)

        # Increased demand due to markdown
        demand_lift = (markdown_pct / 0.5) ** (-demand_elasticity)
        daily_demand_markdown = daily_demand_full * demand_lift

        # Total sales
        sales_full_price = days_full_price * daily_demand_full * regular_price
        sales_markdown = days_markdown * daily_demand_markdown * markdown_price

        # Costs
        units_sold = (days_full_price * daily_demand_full +
                     days_markdown * daily_demand_markdown)
        total_cost = units_sold * cost

        # Profit
        profit = sales_full_price + sales_markdown - total_cost

        # Penalty for waste (unsold inventory)
        # Assume some units don't sell even with markdown
        waste = max(0, 100 - units_sold)  # Assume started with 100 units
        waste_cost = waste * cost

        return profit - waste_cost

    # Optimize markdown day
    result = minimize_scalar(
        lambda x: -expected_profit(x),  # Negative for maximization
        bounds=(0, current_rsl_days),
        method='bounded'
    )

    optimal_day = int(result.x)
    optimal_profit = -result.fun

    # Calculate optimal markdown percentage
    markdown_pct = min(0.5, (current_rsl_days - optimal_day) / current_rsl_days)

    return {
        'optimal_markdown_day': optimal_day,
        'days_until_markdown': optimal_day,
        'markdown_pct': markdown_pct * 100,
        'markdown_price': regular_price * (1 - markdown_pct),
        'expected_profit': optimal_profit
    }


# Example
markdown_strategy = optimize_markdown_timing(
    current_rsl_days=30,
    regular_price=10.00,
    cost=6.00,
    demand_elasticity=-2.0
)

print(f"Start markdown in: {markdown_strategy['days_until_markdown']} days")
print(f"Markdown %: {markdown_strategy['markdown_pct']:.0f}%")
print(f"Markdown Price: ${markdown_strategy['markdown_price']:.2f}")

浪费追踪与分析

class WasteAnalyzer:
    """
    Track and analyze waste from expiration
    """

    def __init__(self):
        self.waste_records = []

    def record_waste(self, waste_data):
        """Record waste event"""
        self.waste_records.append(waste_data)

    def analyze_waste(self):
        """Analyze waste patterns"""

        if not self.waste_records:
            return None

        df = pd.DataFrame(self.waste_records)

        analysis = {
            'total_waste_units': df['quantity'].sum(),
            'total_waste_value': (df['quantity'] * df['unit_cost']).sum(),
            'waste_by_sku': df.groupby('sku').agg({
                'quantity': 'sum',
                'unit_cost': lambda x: (df.loc[x.index, 'quantity'] * x).sum()
            }),
            'waste_by_location': df.groupby('location')['quantity'].sum(),
            'waste_by_reason': df.groupby('reason')['quantity'].sum(),
            'avg_rsl_at_waste': df['rsl_at_waste'].mean()
        }

        # Root cause analysis
        analysis['top_waste_skus'] = analysis['waste_by_sku'].nlargest(10, 'quantity')

        # Calculate waste rate
        if 'total_demand' in df.columns:
            analysis['waste_rate'] = (
                df['quantity'].sum() / df['total_demand'].sum() * 100
            )

        return analysis

    def identify_waste_drivers(self):
        """Identify key drivers of waste"""

        df = pd.DataFrame(self.waste_records)

        drivers = {}

        # 1. Overstocking
        overstock_waste = df[df['reason'] == 'overstock']
        drivers['overstock'] = {
            'waste_pct': len(overstock_waste) / len(df) * 100,
            'value': (overstock_waste['quantity'] * overstock_waste['unit_cost']).sum()
        }

        # 2. Long lead times
        long_lt_waste = df[df['lead_time_days'] > 14]
        drivers['long_lead_time'] = {
            'waste_pct': len(long_lt_waste) / len(df) * 100,
            'value': (long_lt_waste['quantity'] * long_lt_waste['unit_cost']).sum()
        }

        # 3. Poor forecasting
        forecast_error_waste = df[df['forecast_error_pct'].abs() > 20]
        drivers['forecast_error'] = {
            'waste_pct': len(forecast_error_waste) / len(df) * 100,
            'value': (forecast_error_waste['quantity'] *
                     forecast_error_waste['unit_cost']).sum()
        }

        # 4. Improper rotation (should be FEFO but wasn't)
        rotation_waste = df[df['reason'] == 'improper_rotation']
        drivers['improper_rotation'] = {
            'waste_pct': len(rotation_waste) / len(df) * 100,
            'value': (rotation_waste['quantity'] * rotation_waste['unit_cost']).sum()
        }

        return drivers


# Example
analyzer = WasteAnalyzer()

# Record waste events
analyzer.record_waste({
    'date': '2025-01-15',
    'sku': 'SKU_A',
    'location': 'DC1',
    'quantity': 100,
    'unit_cost': 5.00,
    'reason': 'overstock',
    'rsl_at_waste': 0,
    'lead_time_days': 21,
    'forecast_error_pct': 35,
    'total_demand': 500
})

analysis = analyzer.analyze_waste()
drivers = analyzer.identify_waste_drivers()

print(f"Total Waste Value: ${analysis['total_waste_value']:,.0f}")
print(f"Waste Rate: {analysis.get('waste_rate', 0):.1f}%")
print("\nWaste Drivers:")
for driver, data in drivers.items():
    print(f"  {driver}: {data['waste_pct']:.0f}% of waste, ${data['value']:,.0f}")

---

新鲜度优化

基于商品龄期选择供应商

def select_supplier_by_freshness(suppliers, demand, min_rsl_required):
    """
    Select suppliers to maximize freshness

    Parameters:
    - suppliers: list of suppliers with available product and RSL
    - demand: total demand to fulfill
    - min_rsl_required: minimum RSL acceptable

    Returns:
    - optimal supplier selection
    """

    from pulp import *

    # Create problem
    prob = LpProblem("Freshness_Optimization", LpMaximize)

    # Decision variables: quantity from each supplier
    x = LpVariable.dicts("Quantity",
                          [s['supplier_id'] for s in suppliers],
                          lowBound=0,
                          cat='Continuous')

    # Objective: Maximize weighted freshness
    # Higher RSL = better
    objective = lpSum([
        x[s['supplier_id']] * s['rsl_days']
        for s in suppliers
    ])

    prob += objective

    # Constraints

    # 1. Meet demand
    prob += lpSum([x[s['supplier_id']] for s in suppliers]) >= demand

    # 2. Supplier capacity
    for s in suppliers:
        prob += x[s['supplier_id']] <= s['available_quantity']

    # 3. Minimum RSL
    for s in suppliers:
        if s['rsl_days'] < min_rsl_required:
            prob += x[s['supplier_id']] == 0

    # Solve
    prob.solve(PULP_CBC_CMD(msg=0))

    # Extract results
    results = []
    for s in suppliers:
        qty = x[s['supplier_id']].varValue
        if qty > 0:
            results.append({
                'supplier': s['supplier_id'],
                'quantity': qty,
                'rsl_days': s['rsl_days'],
                'cost': qty * s['unit_cost']
            })

    total_qty = sum(r['quantity'] for r in results)
    weighted_rsl = sum(r['quantity'] * r['rsl_days'] for r in results) / total_qty

    return {
        'allocation': results,
        'total_quantity': total_qty,
        'weighted_avg_rsl': weighted_rsl,
        'total_cost': sum(r['cost'] for r in results)
    }


# Example
suppliers = [
    {
        'supplier_id': 'Supplier_A',
        'available_quantity': 500,
        'rsl_days': 90,
        'unit_cost': 5.00
    },
    {
        'supplier_id': 'Supplier_B',
        'available_quantity': 800,
        'rsl_days': 60,
        'unit_cost': 4.80
    },
    {
        'supplier_id': 'Supplier_C',
        'available_quantity': 400,
        'rsl_days': 120,  # Freshest
        'unit_cost': 5.20
    }
]

result = select_supplier_by_freshness(
    suppliers=suppliers,
    demand=1000,
    min_rsl_required=45
)

print("Supplier Allocation:")
for alloc in result['allocation']:
    print(f"  {alloc['supplier']}: {alloc['quantity']} units, "
          f"RSL: {alloc['rsl_days']} days")

print(f"\nWeighted Avg RSL: {result['weighted_avg_rsl']:.0f} days")

---

法规合规

日期代码管理

class DateCodeManager:
    """
    Manage date codes and regulatory compliance
    """

    def __init__(self, date_format='%Y%m%d'):
        self.date_format = date_format

    def parse_date_code(self, date_code, code_type='manufacturing'):
        """
        Parse date code to datetime

        Common formats:
        - YYYYMMDD: 20250115
        - YYMMDD: 250115
        - Julian: 25015 (year + day of year)
        """

        if len(date_code) == 8:  # YYYYMMDD
            return datetime.strptime(date_code, '%Y%m%d')
        elif len(date_code) == 6:  # YYMMDD
            return datetime.strptime(date_code, '%y%m%d')
        elif len(date_code) == 5:  # Julian YYDDD
            year = int('20' + date_code[:2])
            day_of_year = int(date_code[2:])
            return datetime(year, 1, 1) + timedelta(days=day_of_year - 1)
        else:
            raise ValueError(f"Unknown date code format: {date_code}")

    def validate_date_code(self, date_code, product_type='food'):
        """
        Validate date code meets regulatory requirements

        Requirements vary by region and product type
        """

        try:
            parsed_date = self.parse_date_code(date_code)
        except:
            return {'valid': False, 'reason': 'Invalid date code format'}

        current_date = datetime.now()

        # Check if manufacturing date is not in future
        if parsed_date > current_date:
            return {'valid': False, 'reason': 'Manufacturing date in future'}

        # Check if too old (product-specific)
        max_age_days = {
            'food_fresh': 30,
            'food_frozen': 365,
            'food_shelf_stable': 730,
            'pharma': 1825,  # 5 years typically
            'cosmetics': 730
        }

        age_days = (current_date - parsed_date).days
        max_age = max_age_days.get(product_type, 365)

        if age_days > max_age:
            return {
                'valid': False,
                'reason': f'Product too old: {age_days} days (max: {max_age})'
            }

        return {'valid': True, 'parsed_date': parsed_date, 'age_days': age_days}

    def calculate_expiration_date(self, manufacturing_date, shelf_life_days,
                                    sell_by_buffer_days=0):
        """
        Calculate expiration and sell-by dates

        Parameters:
        - manufacturing_date: when product was made
        - shelf_life_days: total shelf life
        - sell_by_buffer_days: days before expiration to stop selling

        Returns:
        - expiration_date, sell_by_date
        """

        if isinstance(manufacturing_date, str):
            manufacturing_date = self.parse_date_code(manufacturing_date)

        expiration_date = manufacturing_date + timedelta(days=shelf_life_days)
        sell_by_date = expiration_date - timedelta(days=sell_by_buffer_days)

        return {
            'manufacturing_date': manufacturing_date,
            'expiration_date': expiration_date,
            'sell_by_date': sell_by_date,
            'shelf_life_days': shelf_life_days
        }


# Example
manager = DateCodeManager()

# Parse date code
date_info = manager.parse_date_code('20250115')
print(f"Parsed Date: {date_info}")

# Validate
validation = manager.validate_date_code('20250115', product_type='food_shelf_stable')
print(f"Valid: {validation['valid']}")

# Calculate expiration
expiry = manager.calculate_expiration_date(
    manufacturing_date='20250115',
    shelf_life_days=180,
    sell_by_buffer_days=7
)

print(f"Expiration Date: {expiry['expiration_date']}")
print(f"Sell-By Date: {expiry['sell_by_date']}")

---

工具与技术

保质期管理软件

支持 FEFO 的仓库管理系统（WMS）：

• Manhattan Associates WMS：高级 FEFO 和批次追踪
• Blue Yonder WMS：保质期管理
• SAP EWM：支持到期管理的扩展仓库管理
• Oracle WMS：支持日期代码和 FEFO
• HighJump WMS：易腐品管理

专业解决方案：

• FoodLogiQ：食品可追溯性和日期代码管理
• Trace Register：供应链可追溯性
• rfxcel：序列化和到期追踪
• FreshSurety：保质期和温度监控
• ZestIOT：实时新鲜度监控

降价优化：

• Revionics：价格和降价优化（Oracle）
• Pricefx：支持到期管理的动态定价
• PROS：AI 驱动的降价优化

Python 库

# Date handling
from datetime import datetime, timedelta
import pandas as pd
import numpy as np

# Optimization
from pulp import *
from scipy.optimize import minimize, minimize_scalar

# Machine learning for forecasting
from sklearn.ensemble import RandomForestRegressor
from sklearn.linear_model import LinearRegression

# Data visualization
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px

---

常见挑战与解决方案

挑战：浪费率高

问题：

• 5-10% 的库存过期
• 造成显著的成本影响
• 收入损失

解决方案：

• 严格实施 FEFO
• 减少订购量（提高订购频率）
• 改进需求预测
• 动态调整安全库存（随着临近到期而减少）
• 提前并更大幅度地降价
• 捐赠临近到期的产品（可获得税收优惠，提升商誉）

挑战：日期代码格式不一致

问题：

• 供应商使用不同格式
• 人工追踪容易出错
• 存在合规风险

解决方案：

• 统一供应商使用的日期代码格式
• 自动解析日期代码（OCR、条码）
• 收货时进行验证
• 与 WMS 集成
• 管理主数据

挑战：客户新鲜度要求

问题：

• 零售商要求最低 RSL 达到 75%
• 限制可用库存
• 增加配送中心（DC）的浪费

解决方案：

• 协商 RSL 要求
• 为较低的 RSL 提供价格激励
• 快速向门店补货
• 为要求严格的客户分配更新鲜的库存
• 将较旧的库存用于促销

挑战：多级网络复杂性

问题：

• DC 持有日渐陈旧的库存
• 门店也有新鲜度要求
• 难以优化整个网络

解决方案：

• 全网络范围的 RSL 可视性
• 集中分配（将最新鲜的库存分配到最远的地点）
• 根据有效期动态调整路线
• 为快速流转商品采用越库作业
• 新鲜商品绕过 DC

---

输出格式

保质期绩效报告

执行摘要：

• 总库存：500,000 件
• 浪费率：3.2%（去年为 5.1%）
• 浪费金额：每年 $320,000
• 销售时的平均 RSL：68%
• 合规率：100%（未售出过期商品）

过期情况摘要：

区域	件数	占总量比例	所需操作
绿色（RSL >67%）	350,000	70%	正常销售
黄色（RSL 为 33-67%）	100,000	20%	优先出库
红色（RSL 为 10-33%）	45,000	9%	立即降价
已过期（RSL <10%）	5,000	1%	立即撤下

未来 30 天内即将过期：

SKU	位置	数量	到期日期	RSL 天数	操作
SKU_A	DC1	2,500	2025-02-15	15	降价 30%
SKU_B	DC2	1,200	2025-02-10	10	降价 50%
SKU_C	DC1	800	2025-02-05	5	撤下/捐赠

浪费分析：

类别	浪费件数	金额	占浪费总量比例
库存过剩	8,000	$160,000	50%
预测误差	4,000	$80,000	25%
前置时间过长	3,000	$60,000	18.75%
轮转不当	1,000	$20,000	6.25%

建议：

1. 实施自动化 FEFO 分配（减少轮转错误）
2. 减少 SKU_A、SKU_B 的订货量（高浪费商品）
3. 提前触发滞销商品降价（红色区域 → RSL 达到 40% 时降价）
4. 与食品银行合作开展捐赠计划
5. 与零售商协商延长 RSL 要求

---

需要询问的问题

如需更多背景信息：

1. 哪些产品有保质期方面的顾虑？保质期多长？
2. 当前浪费/变质率和成本是多少？
3. WMS 是否支持 FEFO？
4. 客户的 RSL 要求是什么？
5. 日期代码的跟踪方式和格式是什么？
6. 降价流程和时间安排是什么？
7. 是多级网络还是单一地点？
8. 有哪些监管要求（FDA、USDA 等）？

---

相关技能

• inventory-optimization：用于考虑有效期限制的安全库存管理
• demand-forecasting：用于减少库存过剩和浪费
• warehouse-slotting-optimization：用于支持 FEFO 的库位规划
• food-beverage-supply-chain：用于易腐产品供应链
• pharmaceutical-supply-chain：用于药品有效期管理
• markdown-optimization：用于优化即将过期产品的价格
• quality-management：用于质量控制和合规
• replenishment-strategy：用于制定考虑有效期的最优补货策略