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

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Last-Mile Route Optimisation Agent

Last-Mile Route Optimisation Agent

Author: Venkata Sudhakar

Last-mile delivery is the most expensive and time-consuming leg of the supply chain. For ShopMax India, optimising the sequence of stops for each delivery vehicle can cut fuel costs, reduce driver hours, and improve on-time rates across Mumbai, Bangalore, and Delhi.

This tutorial builds a Gemini ADK agent that groups pending orders by delivery zone and computes an optimised stop sequence using a nearest-neighbour heuristic. The agent reports estimated distance, ETA, and load per vehicle.

The below example shows a route optimisation 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

# Pending orders with lat/lon and weight
ORDERS = [
    {"id": "ORD-201", "lat": 19.076, "lon": 72.877, "kg": 2.5, "zone": "A"},
    {"id": "ORD-202", "lat": 19.083, "lon": 72.881, "kg": 1.2, "zone": "A"},
    {"id": "ORD-203", "lat": 19.071, "lon": 72.869, "kg": 3.0, "zone": "A"},
    {"id": "ORD-204", "lat": 19.120, "lon": 72.905, "kg": 4.0, "zone": "B"},
    {"id": "ORD-205", "lat": 19.115, "lon": 72.910, "kg": 1.8, "zone": "B"},
    {"id": "ORD-206", "lat": 19.055, "lon": 72.855, "kg": 2.2, "zone": "C"},
]

DEPOT = {"lat": 19.060, "lon": 72.860}  # ShopMax Mumbai warehouse
VEHICLE_CAPACITY_KG = 10.0

def haversine(a, b):
    R = 6371
    dlat = math.radians(b["lat"] - a["lat"])
    dlon = math.radians(b["lon"] - a["lon"])
    h = (math.sin(dlat/2)**2 +
         math.cos(math.radians(a["lat"])) *
         math.cos(math.radians(b["lat"])) *
         math.sin(dlon/2)**2)
    return 2 * R * math.asin(math.sqrt(h))

def nearest_neighbour_route(stops):
    unvisited = list(stops)
    route = []
    current = DEPOT
    while unvisited:
        nearest = min(unvisited, key=lambda s: haversine(current, s))
        route.append(nearest)
        current = nearest
        unvisited.remove(nearest)
    return route

def optimise_routes(zone: str) -> dict:
    zone_orders = [o for o in ORDERS if o["zone"] == zone]
    if not zone_orders:
        return {"error": f"No orders found for zone {zone}"}
    total_kg = sum(o["kg"] for o in zone_orders)
    vehicles_needed = math.ceil(total_kg / VEHICLE_CAPACITY_KG)
    route = nearest_neighbour_route(zone_orders)
    dist = haversine(DEPOT, route[0])
    for i in range(len(route) - 1):
        dist += haversine(route[i], route[i+1])
    dist += haversine(route[-1], DEPOT)
    eta_min = int(dist / 30 * 60)  # assume 30 km/h avg
    return {
        "zone": zone,
        "orders": len(zone_orders),
        "total_kg": round(total_kg, 1),
        "vehicles_needed": vehicles_needed,
        "stop_sequence": [o["id"] for o in route],
        "total_distance_km": round(dist, 2),
        "estimated_minutes": eta_min,
    }

route_tool = FunctionTool(func=optimise_routes)

agent = LlmAgent(
    name="route_optimiser",
    model="gemini-2.0-flash",
    description="Optimises last-mile delivery routes by zone for ShopMax India.",
    instruction=(
        "You are a logistics agent for ShopMax India. "
        "When asked to optimise routes for a delivery zone, call optimise_routes with the zone letter. "
        "Present the stop sequence, total distance, estimated time, and vehicle count clearly."
    ),
    tools=[route_tool],
)

async def main():
    session_service = InMemorySessionService()
    runner = Runner(agent=agent, app_name="route_app", session_service=session_service)
    session = await session_service.create_session(app_name="route_app", user_id="ops1")
    msg = types.Content(role="user", parts=[types.Part(text="Optimise delivery routes for zone A")])
    async for event in runner.run_async(user_id="ops1", 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,

Zone A Delivery Route - ShopMax Mumbai

Orders: 3  |  Total Load: 6.7 kg  |  Vehicles: 1

Optimised Stop Sequence:
  1. ORD-203  (nearest to depot)
  2. ORD-201
  3. ORD-202
  -> Return to depot

Total Distance : 4.82 km
Estimated Time : 9 minutes

For ShopMax India, running this agent each morning groups the day's orders by zone, assigns vehicles based on load, and gives drivers a fuel-efficient stop sequence. Integrating real-time traffic data via the Google Maps API or a traffic-aware routing service further reduces delivery time and fuel costs.

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