	In Browser
	StumbleUpon
	del.icio.us
	Google
	Google Buzz
	reddit
	LinkedIn

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Packaging Optimisation Agent

Packaging Optimisation Agent

Author: Venkata Sudhakar

Shipping carriers charge by dimensional weight, which is calculated from box volume. Selecting an oversized box wastes material and inflates shipping costs. For ShopMax India, which ships thousands of electronics orders daily, even small improvements in box selection yield significant savings.

This tutorial builds a Gemini ADK agent that accepts order item dimensions and weights, selects the most space-efficient standard box from ShopMax's packaging catalogue, and reports void fill required and estimated shipping weight.

The below example shows a packaging selection agent for ShopMax India in a business context.

import asyncio
import math
from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService
from google.genai import types

# Standard box catalogue: name, (L x W x H in cm), max_weight_kg
BOXES = [
    {"name": "XS", "l": 20, "w": 15, "h": 10, "max_kg": 2},
    {"name": "S",  "l": 30, "w": 20, "h": 15, "max_kg": 5},
    {"name": "M",  "l": 40, "w": 30, "h": 20, "max_kg": 10},
    {"name": "L",  "l": 50, "w": 40, "h": 30, "max_kg": 20},
    {"name": "XL", "l": 70, "w": 50, "h": 40, "max_kg": 30},
]

DIM_DIVISOR = 5000  # carrier dimensional weight divisor (cm^3 / kg)

def select_box(item_l: float, item_w: float, item_h: float, item_weight_kg: float) -> dict:
    # Add 3 cm clearance on each dimension for cushioning
    needed_l = item_l + 3
    needed_w = item_w + 3
    needed_h = item_h + 3
    candidates = [
        b for b in BOXES
        if b["l"] >= needed_l and b["w"] >= needed_w and
           b["h"] >= needed_h and b["max_kg"] >= item_weight_kg
    ]
    if not candidates:
        return {"error": "No standard box fits this item. Custom packaging required."}
    best = min(candidates, key=lambda b: b["l"] * b["w"] * b["h"])
    box_vol = best["l"] * best["w"] * best["h"]
    item_vol = item_l * item_w * item_h
    void_pct = round((1 - item_vol / box_vol) * 100, 1)
    dim_weight = round(box_vol / DIM_DIVISOR, 2)
    chargeable = max(item_weight_kg, dim_weight)
    return {
        "box": best["name"],
        "box_dims_cm": f"{best['l']}x{best['w']}x{best['h']}",
        "item_dims_cm": f"{item_l}x{item_w}x{item_h}",
        "void_fill_pct": void_pct,
        "actual_weight_kg": item_weight_kg,
        "dim_weight_kg": dim_weight,
        "chargeable_weight_kg": chargeable,
    }

box_tool = FunctionTool(func=select_box)

agent = LlmAgent(
    name="packaging_agent",
    model="gemini-2.0-flash",
    description="Recommends optimal box size for ShopMax India orders to minimise shipping cost.",
    instruction=(
        "You are a packaging optimisation agent for ShopMax India. "
        "When given item dimensions (L, W, H in cm) and weight (kg), "
        "call select_box and present the recommended box, void fill percentage, "
        "and chargeable shipping weight clearly."
    ),
    tools=[box_tool],
)

async def main():
    session_service = InMemorySessionService()
    runner = Runner(agent=agent, app_name="pack_app", session_service=session_service)
    session = await session_service.create_session(app_name="pack_app", user_id="wh1")
    msg = types.Content(role="user", parts=[types.Part(
        text="Item: Sony headphones, 25x18x12 cm, 0.9 kg. Which box should we use?"
    )])
    async for event in runner.run_async(user_id="wh1", session_id=session.id, new_message=msg):
        if event.is_final_response() and event.content:
            print(event.content.parts[0].text)

asyncio.run(main())

It gives the following output,

Packaging Recommendation - ShopMax India

Product       : Sony Headphones
Item Dims     : 25 x 18 x 12 cm
Recommended Box: S (30 x 20 x 15 cm)

Void Fill     : 37.5%  (cushioning material needed)
Actual Weight : 0.9 kg
Dim Weight    : 0.18 kg
Chargeable    : 0.9 kg  (actual wins)

Box S is the smallest standard box that fits with safety clearance.

ShopMax India can run this agent at the picking station - warehouse staff scan the product, the agent recommends the right box instantly. Over a month, selecting correctly-sized boxes reduces void fill material cost and eliminates dimensional weight surcharges on lightweight but bulky items such as headphone boxes and laptop bags.

Send your comments, suggestions or queries regarding this site to [email protected].