submission

README.md

@@ -0,0 +1,131 @@
+# Elevator Simulation
+
+## overview
+
+the code models an elevator inspired by event-driven entities in game development systems.
+
+the goal here is to simulate an elevator and to use this simulation to generate data to train an ML model.
+
+## how it works
+
+the code is split into the following:
+
+```
+devtest/
+├── models/
+│   ├── elevator.py      # core elevator logic
+│   └── building.py      # manages the building and elevators
+├── simulation/
+│   └── simulator.py     # main simulation runner
+├── data/
+│   └── collector.py     # saves data to sqlite
+├── tests/
+│   └── test_elevator.py # simple tests that check if system is okay
+├── config.py            # settings for the simulation
+└── main.py              # demos
+```
+
+the simulation runs in discrete time steps (called _ticks_, inspired by game engines). in each tick, the elevator can move one floor.
+
+after taking a quick look online, i found out that elevators generally use the SCAN algorithm: the elevator keeps going in one direction as long as there are requests in that direction. once it's done, it switches directions if needed. this approach is more efficient than jumping around for every new request.
+
+we also have a couple of "business rules":
+- if an elevator is 80% full, it won't accept new hall calls.
+- if an elevator has been moving for 5 minutes straight, it takes a 30-second maintenance break.
+
+both of these rules are parameterized, so their values can be changed.
+
+## how to use it
+
+### run the demo
+
+to see it in action, just run:
+```bash
+python main.py
+```
+
+here's the terminal output:
+```
+=== ELEVATOR SIMULATION DEMO ===
+running test scenario:
+- request floor 3 (elevator should go UP)
+- request floor 2 while moving up (should serve 2 then 3)
+- request floor 1 and 0 (should serve both going DOWN)
+running scenario with 4 requests...
+request for floor 3: success
+all elevators idle, continuing to next request...
+request for floor 2: success
+all elevators idle, continuing to next request...
+request for floor 1: success
+all elevators idle, continuing to next request...
+request for floor 0: success
+all elevators idle, continuing to next request...
+scenario completed in 9.0 seconds
+scenario generated 10 events
+final elevator state:
+- current floor: 0
+- direction: IDLE
+- pending requests: []
+- time moving: 2.0s
+- time idle: 4.0s
+- occupancy: 0/8
+- near capacity: False
+- maintenance mode: False
+
+=== RANDOM SIMULATION FOR DATA COLLECTION ===
+starting simulation for 60 seconds...
+added request for floor 3
+elevator elevator_1: arrived at floor 3
+added request for floor 5
+elevator elevator_1: arrived at floor 5
+added request for floor 2
+elevator elevator_1: arrived at floor 2
+added request for floor 5
+elevator elevator_1: arrived at floor 5
+added request for floor 3
+elevator elevator_1: arrived at floor 3
+added request for floor 5
+elevator elevator_1: arrived at floor 5
+added request for floor 3
+elevator elevator_1: arrived at floor 3
+simulation completed after 60 ticks
+collected 34 events
+data summary:
+- total events recorded: 34
+- simulation duration: 60.3s
+- storage type: sqlite
+- df shape: (34, 12)
+- event types: {'moving': 23, 'arrived': 11}
+- data exported to: elevator_simulation_data.csv
+simulation complete!
+```
+
+### run tests
+to run the tests:
+```bash
+python tests/test_elevator.py
+```
+
+### configuration
+
+you can change things like the number of floors and elevators in `config.py`.
+
+## the data
+
+all the simulation events are saved in a sqlite database file called `elevator_simulation.db`. this makes it easy to use the data later.
+
+here's what we save for each event:
+- `timestamp`: time of the event
+- `elevator_id`: id of the elevator
+- `event_type`: what happened (e.g., `moving`, `arrived`, `became_idle`)
+- `current_floor`: where the elevator is
+- `direction`: which way the elevator is going
+- `pending_requests`: how many requests are left
+- `occupancy`: how many people are inside
+- `time_since_last_request`: time since the last button was pressed
+- `hour_of_day`: hour of day when the event happened
+- `day_of_week`: day of week when the event happened
+
+### exporting to csv
+
+after running a simulation, the data is exported a CSV file, which allows it to be easily fed into an ML model via `pandas`.

config.py

@@ -0,0 +1,20 @@
+# building setup
+BUILDING_CONFIG = {
+    'floors': [0, 1, 2, 3, 4, 5], # available floors (0 is ground level)
+    'num_elevators': 1, # number of elevators
+    'elevator_capacity': 8, # max people that fit inside
+}
+
+# timing settings
+SIMULATION_CONFIG = {
+    'tick_duration': 1.0, # how long each tick lasts (seconds)
+    'floor_travel_time': 3.0, # time to move between floors
+    'door_time': 2.0,  # time for doors to open/close
+    'max_wait_time': 60.0, # how long elevator waits around doing nothing
+}
+
+# where we save all the data for ML training
+DATA_CONFIG = {
+    'db_file': 'elevator_simulation.db', # sqlite database file
+    'collect_events': True,
+} 

data/collector.py

@@ -0,0 +1,172 @@
+import time
+from datetime import datetime
+from typing import List, Dict, Any
+import sqlite3
+import pandas as pd
+
+
+class SQLiteDataCollector:
+    """
+    collects all the elevator events and saves them to sqlite
+    """
+
+    def __init__(self, db_file: str = "elevator_simulation.db"):
+        self.db_file = db_file
+        self.start_time = time.time()
+        self.last_request_time = self.start_time
+
+        self._initialize_database()
+
+    def _get_connection(self):
+        """
+        connect to the sqlite db
+        """
+        return sqlite3.connect(self.db_file)
+
+    def _initialize_database(self):
+        """
+        set up the db table
+        """
+        create_table_sql = """
+        CREATE TABLE IF NOT EXISTS elevator_events (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            timestamp REAL NOT NULL,
+            simulation_time REAL NOT NULL,
+            elevator_id TEXT NOT NULL,
+            event_type TEXT NOT NULL,
+            current_floor INTEGER NOT NULL,
+            direction INTEGER NOT NULL,
+            is_moving BOOLEAN NOT NULL,
+            occupancy INTEGER NOT NULL,
+            pending_requests INTEGER NOT NULL,
+            time_since_last_request REAL NOT NULL,
+            hour_of_day INTEGER NOT NULL,
+            day_of_week INTEGER NOT NULL,
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        );
+        """
+
+        with self._get_connection() as conn:
+            cursor = conn.cursor()
+            cursor.execute(create_table_sql)
+            conn.commit()
+
+    def record_events(self, events: List[Dict[str, Any]], building_status: Dict):
+        """
+        save elevator events with extra info for ML training
+        """
+        if not events:
+            return
+
+        current_time = time.time()
+        simulation_time = current_time - self.start_time
+
+        # get time features
+        sim_datetime = datetime.fromtimestamp(current_time)
+        hour_of_day = sim_datetime.hour
+        day_of_week = sim_datetime.weekday()
+
+        rows = []
+
+        for event in events:
+            # get elevator info
+            elevator_id = event.get('elevator_id', 'unknown')
+            elevator_info = building_status.get('elevators', {}).get(elevator_id, {})
+
+            # count how many requests are still pending
+            pending_requests = (
+                len(elevator_info.get('up_requests', [])) + 
+                len(elevator_info.get('down_requests', []))
+            )
+
+            time_since_last_request = current_time - self.last_request_time
+
+            # write everything into a row for the db
+            row_data = (
+                current_time,
+                simulation_time,
+                elevator_id,
+                event.get('type', 'unknown'),
+                event.get('floor', elevator_info.get('current_floor', 0)),
+                event.get('direction', 0),
+                elevator_info.get('is_moving', False),
+                elevator_info.get('occupancy', 0),
+                pending_requests,
+                time_since_last_request,
+                hour_of_day,
+                day_of_week,
+            )
+
+            rows.append(row_data)
+
+        self._insert_rows(rows)
+
+    def _insert_rows(self, rows: List[tuple]):
+        """
+        insert rows into db
+        """
+        insert_sql = """
+        INSERT INTO elevator_events (
+            timestamp, simulation_time, elevator_id, event_type, current_floor,
+            direction, is_moving, occupancy, pending_requests, time_since_last_request,
+            hour_of_day, day_of_week
+        ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+        """
+
+        with self._get_connection() as conn:
+            cursor = conn.cursor()
+            cursor.executemany(insert_sql, rows)
+            conn.commit()
+
+    def record_request(self, floor: int, elevator_id: str):
+        """
+        get last request time
+        """
+        self.last_request_time = time.time()
+
+    def get_data_summary(self) -> Dict:
+        """
+        get a quick summary of how much data was collected
+        """
+        with self._get_connection() as conn:
+            cursor = conn.cursor()
+            cursor.execute("SELECT COUNT(*) FROM elevator_events")
+            event_count = cursor.fetchone()[0]
+
+        return {
+            'total_events': event_count,
+            'storage_type': 'sqlite',
+            'simulation_duration': time.time() - self.start_time
+        }
+
+    def get_dataframe(self) -> pd.DataFrame:
+        """
+        get all the data as a pandas dataframe
+        """
+        query = """
+        SELECT timestamp, simulation_time, elevator_id, event_type, current_floor,
+               direction, is_moving, occupancy, pending_requests, time_since_last_request,
+               hour_of_day, day_of_week
+        FROM elevator_events 
+        ORDER BY timestamp
+        """
+
+        with self._get_connection() as conn:
+            return pd.read_sql_query(query, conn)
+
+    def export_to_csv(self, filename: str = "elevator_data.csv") -> str:
+        """
+        export all data to a csv file
+        """
+        df = self.get_dataframe()
+        df.to_csv(filename, index=False)
+        return filename
+
+    def clear_data(self):
+        """
+        clear all data from the database (useful for testing)
+        """
+        with self._get_connection() as conn:
+            cursor = conn.cursor()
+            cursor.execute("DELETE FROM elevator_events")
+            conn.commit() 

elevator_simulation_data.csv

@@ -0,0 +1,35 @@
+timestamp,simulation_time,elevator_id,event_type,current_floor,direction,is_moving,occupancy,pending_requests,time_since_last_request,hour_of_day,day_of_week
+1751331166.891501,0.0009388923645019531,elevator_1,moving,1,1,1,0,1,1.5974044799804688e-05,21,0
+1751331167.897501,1.0069389343261719,elevator_1,moving,2,1,1,0,1,1.0060160160064697,21,0
+1751331168.903875,2.013313055038452,elevator_1,moving,3,1,0,0,0,2.01239013671875,21,0
+1751331168.903875,2.013313055038452,elevator_1,arrived,3,0,0,0,0,2.01239013671875,21,0
+1751331170.911757,4.021194934844971,elevator_1,moving,2,-1,0,0,0,0.0003178119659423828,21,0
+1751331170.911757,4.021194934844971,elevator_1,arrived,2,0,0,0,0,0.0003178119659423828,21,0
+1751331173.922013,7.031450986862183,elevator_1,moving,1,-1,0,0,0,0.0003349781036376953,21,0
+1751331173.922013,7.031450986862183,elevator_1,arrived,1,0,0,0,0,0.0003349781036376953,21,0
+1751331175.9330041,9.042442083358765,elevator_1,moving,0,-1,0,0,0,0.00030112266540527344,21,0
+1751331175.9330041,9.042442083358765,elevator_1,arrived,0,0,0,0,0,0.00030112266540527344,21,0
+1751331183.9714549,8.033185958862305,elevator_1,moving,1,1,1,0,1,0.0002639293670654297,21,0
+1751331184.980256,9.041987180709839,elevator_1,moving,2,1,1,0,1,1.0090651512145996,21,0
+1751331185.98958,10.051311016082764,elevator_1,moving,3,1,0,0,0,2.0183889865875244,21,0
+1751331185.98958,10.051311016082764,elevator_1,arrived,3,0,0,0,0,2.0183889865875244,21,0
+1751331192.0190668,16.080797910690308,elevator_1,moving,4,1,1,0,1,0.0003058910369873047,21,0
+1751331193.0296419,17.091372966766357,elevator_1,moving,5,1,0,0,0,1.010880947113037,21,0
+1751331193.0296419,17.091372966766357,elevator_1,arrived,5,0,0,0,0,1.010880947113037,21,0
+1751331200.066695,24.12842607498169,elevator_1,moving,4,-1,1,0,1,0.000370025634765625,21,0
+1751331201.076561,25.13829207420349,elevator_1,moving,3,-1,1,0,1,1.0102360248565674,21,0
+1751331202.085664,26.147395133972168,elevator_1,moving,2,-1,0,0,0,2.019339084625244,21,0
+1751331202.085664,26.147395133972168,elevator_1,arrived,2,0,0,0,0,2.019339084625244,21,0
+1751331208.1116312,32.173362255096436,elevator_1,moving,3,1,1,0,1,0.0002541542053222656,21,0
+1751331209.119939,33.18167018890381,elevator_1,moving,4,1,1,0,1,1.0085620880126953,21,0
+1751331210.130243,34.19197416305542,elevator_1,moving,5,1,0,0,0,2.0188660621643066,21,0
+1751331210.130243,34.19197416305542,elevator_1,arrived,5,0,0,0,0,2.0188660621643066,21,0
+1751331216.157873,40.21960401535034,elevator_1,moving,4,-1,1,0,1,6.794929504394531e-05,21,0
+1751331217.16702,41.22875118255615,elevator_1,moving,3,-1,0,0,0,1.0092151165008545,21,0
+1751331217.16702,41.22875118255615,elevator_1,arrived,3,0,0,0,0,1.0092151165008545,21,0
+1751331224.1976562,48.259387254714966,elevator_1,moving,4,1,1,0,1,0.0002770423889160156,21,0
+1751331225.207409,49.269140005111694,elevator_1,moving,5,1,0,0,0,1.0100297927856445,21,0
+1751331225.207409,49.269140005111694,elevator_1,arrived,5,0,0,0,0,1.0100297927856445,21,0
+1751331232.243977,56.305708169937134,elevator_1,moving,4,-1,1,0,1,0.0002791881561279297,21,0
+1751331233.2546718,57.31640291213989,elevator_1,moving,3,-1,0,0,0,1.0109739303588867,21,0
+1751331233.2546718,57.31640291213989,elevator_1,arrived,3,0,0,0,0,1.0109739303588867,21,0