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

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Cross-Docking Agent

Cross-Docking Agent

Author: Venkata Sudhakar

Cross-docking eliminates warehouse storage by transferring goods directly from inbound trucks to outbound delivery vehicles. For ShopMax India's regional distribution hubs, this cuts handling time and storage costs while speeding up same-day fulfilment for metro orders.

This tutorial builds a Gemini ADK agent that receives an inbound manifest, matches each order to an available outbound vehicle by destination zone and capacity, and flags any unmatched items needing temporary staging.

The below example shows a cross-docking coordination agent for ShopMax India in a business context.

import asyncio
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

# Inbound shipment manifest arriving at hub
INBOUND = [
    {"id": "IN-001", "zone": "North Mumbai", "kg": 3.5, "orders": ["ORD-701", "ORD-702"]},
    {"id": "IN-002", "zone": "South Mumbai", "kg": 6.0, "orders": ["ORD-703", "ORD-704", "ORD-705"]},
    {"id": "IN-003", "zone": "Thane",        "kg": 2.0, "orders": ["ORD-706"]},
    {"id": "IN-004", "zone": "Navi Mumbai",  "kg": 8.5, "orders": ["ORD-707", "ORD-708"]},
]

# Outbound vehicles docked and ready
OUTBOUND_VEHICLES = [
    {"id": "VH-A", "zone": "North Mumbai", "capacity_kg": 10.0, "loaded_kg": 2.0},
    {"id": "VH-B", "zone": "South Mumbai", "capacity_kg": 15.0, "loaded_kg": 5.0},
    {"id": "VH-C", "zone": "Thane",        "capacity_kg": 8.0,  "loaded_kg": 0.0},
]

def run_cross_dock() -> dict:
    vehicles = {v["id"]: dict(v) for v in OUTBOUND_VEHICLES}
    matched = []
    staged = []
    for item in INBOUND:
        # Find a vehicle for the same zone with enough remaining capacity
        target = None
        for v in vehicles.values():
            if v["zone"] == item["zone"]:
                remaining = v["capacity_kg"] - v["loaded_kg"]
                if remaining >= item["kg"]:
                    target = v
                    break
        if target:
            target["loaded_kg"] += item["kg"]
            matched.append({
                "inbound_id": item["id"],
                "zone": item["zone"],
                "orders": item["orders"],
                "kg": item["kg"],
                "assigned_vehicle": target["id"],
                "vehicle_fill_pct": round(target["loaded_kg"] / target["capacity_kg"] * 100, 1),
            })
        else:
            staged.append({
                "inbound_id": item["id"],
                "zone": item["zone"],
                "orders": item["orders"],
                "kg": item["kg"],
                "reason": "No vehicle available for zone or insufficient capacity",
            })
    return {
        "matched": len(matched),
        "staged": len(staged),
        "assignments": matched,
        "staging_required": staged,
    }

dock_tool = FunctionTool(func=run_cross_dock)

agent = LlmAgent(
    name="cross_dock_agent",
    model="gemini-2.0-flash",
    description="Coordinates cross-docking at ShopMax India hub, matching inbound to outbound vehicles.",
    instruction=(
        "You are a cross-docking coordinator for ShopMax India. "
        "When asked to process the inbound manifest, call run_cross_dock. "
        "Report each matched assignment with vehicle ID and fill percentage, "
        "then list any items that need temporary staging and why."
    ),
    tools=[dock_tool],
)

async def main():
    session_service = InMemorySessionService()
    runner = Runner(agent=agent, app_name="dock_app", session_service=session_service)
    session = await session_service.create_session(app_name="dock_app", user_id="hub1")
    msg = types.Content(role="user", parts=[types.Part(
        text="Process the inbound manifest and assign vehicles for cross-docking"
    )])
    async for event in runner.run_async(user_id="hub1", 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,

Cross-Dock Assignment Report - ShopMax Mumbai Hub

Matched : 3 of 4 inbound items
Staged  : 1 item requires temporary storage

Assignments:
  IN-001 -> VH-A (North Mumbai)  | 3.5 kg | Vehicle fill: 55.0%
  IN-002 -> VH-B (South Mumbai)  | 6.0 kg | Vehicle fill: 73.3%
  IN-003 -> VH-C (Thane)         | 2.0 kg | Vehicle fill: 25.0%

Staging Required:
  IN-004 - Navi Mumbai | 8.5 kg | Orders: ORD-707, ORD-708
  Reason: No vehicle available for zone or insufficient capacity
  Action: Dispatch additional vehicle VH-D to Navi Mumbai zone

ShopMax India can trigger this agent automatically when inbound trucks check in at the hub gate. The agent's output feeds the warehouse management system to print dock door assignments for each driver, cutting manual coordination time from 20 minutes to under 2 minutes per truck arrival.

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