3 Capstone Projects: Selection and Fleet Tracking

3.9.1.5 System Architecture

Data and command flow:

Step 1 — Device wakes based on movement state or scheduled reporting interval. Step 2 — Firmware obtains a GPS fix, packages location and battery status, and transmits over LTE-M or NB-IoT. Step 3 — Backend stores the point, checks geofence transitions, and updates the dashboard. Step 4 — Operators review alerts, routes, and device health from the web interface.

3.9.1.6 Power Optimization Strategy

// Power states for maximum battery life
enum PowerState {
    DEEP_SLEEP,      // GPS off, cellular off, ~10uA
    LIGHT_SLEEP,     // GPS acquiring, cellular off, ~1mA
    ACTIVE_GPS,      // GPS tracking, cellular off, ~30mA
    ACTIVE_TRANSMIT  // GPS + cellular active, ~150mA
};

// Adaptive reporting based on movement
void adaptiveTracking() {
    if (isMoving()) {
        // Vehicle in motion: frequent updates
        reportIntervalSeconds = 60;
        gpsMode = CONTINUOUS;
    } else if (hasMovedRecently(30)) {
        // Recently stopped: medium updates
        reportIntervalSeconds = 300;
        gpsMode = PERIODIC;
    } else {
        // Parked: infrequent updates
        reportIntervalSeconds = 3600;
        gpsMode = ON_DEMAND;
    }
}

// Estimated battery life calculation
// 3000mAh battery
// Moving 4 hours/day: ~35 days
// Parked most of time: ~90+ days

3.9.1.7 Implementation Milestones

3.9.1.8 Geofence Implementation

from shapely.geometry import Point, Polygon

class GeofenceEngine:
    def __init__(self):
        self.geofences = {}
        self.device_states = {}  # Track inside/outside state

    def add_geofence(self, name, coordinates, alert_on):
        """
        coordinates: List of (lat, lon) tuples
        alert_on: 'entry', 'exit', or 'both'
        """
        self.geofences[name] = {
            'polygon': Polygon(coordinates),
            'alert_on': alert_on
        }

    def check_position(self, device_id, lat, lon):
        """Check if device triggers any geofence alerts."""
        point = Point(lon, lat)  # Note: Shapely uses (x, y) = (lon, lat)
        alerts = []

        for name, geofence in self.geofences.items():
            was_inside = self.device_states.get((device_id, name), None)
            is_inside = geofence['polygon'].contains(point)

            if was_inside is not None:
                if not was_inside and is_inside and geofence['alert_on'] in ('entry', 'both'):
                    alerts.append({'type': 'entry', 'geofence': name})
                elif was_inside and not is_inside and geofence['alert_on'] in ('exit', 'both'):
                    alerts.append({'type': 'exit', 'geofence': name})

            self.device_states[(device_id, name)] = is_inside

        return alerts

3.9.1.9 Evaluation Rubric

Read the rubric as one end-to-end tracking claim. GPS accuracy and cellular delivery establish trustworthy observations, battery life determines how long that evidence can be collected, geofencing and dashboards turn it into operational decisions, and documentation makes every result reproducible. A strong score therefore depends on the connections between criteria, not isolated demonstrations.