You’ve probably checked package tracking dozens of times, watching as your shipment moves from city to sorting facility to “out for delivery.” But have you ever wondered how that tracking actually works?

The technology behind real-time package tracking is more complex and fascinating than most people realize. Let’s pull back the curtain on how logistics companies know where your package is at any given moment—and why that information isn’t always as real-time as we’d like.

The Foundation: Barcodes and Scanning

Everything starts with a barcode or QR code on your package. When a courier first accepts your shipment, they scan it, creating a digital record in the logistics company’s database. That scan associates the barcode with your tracking number, destination, weight, sender details, and other metadata.

As the package moves through the logistics network, it gets scanned repeatedly. Arriving at a sorting facility? Scan. Loading onto a truck? Scan. Arriving at the destination city? Scan. Each scan updates the tracking database with location and timestamp information.

This seems straightforward, but consistency is the challenge. Missed scans create tracking gaps. A package might physically be on a delivery truck but still show as “processing at sorting facility” because someone forgot to scan it during loading.

Indonesia’s logistics companies have gotten much better at scan discipline over the past few years. JNE, J&T, and SiCepat have particularly robust scanning protocols. But occasional gaps still occur, especially during high-volume periods like Ramadan or Harbolnas (Indonesia’s online shopping festival).

GPS Tracking for Vehicles

Many delivery vehicles now have GPS trackers that continuously report their location. This allows logistics companies to know where their trucks are in real-time, even between scanning checkpoints.

When you see a map showing your package’s current location while “out for delivery,” you’re typically seeing the GPS position of the delivery vehicle it’s assigned to. The system knows your package was scanned onto that vehicle and the vehicle is at a specific location, so it infers your package is there too.

This works well until it doesn’t. Sometimes packages get transferred between vehicles without proper scanning. Or a courier might leave packages at a hub while taking the vehicle out for other deliveries. The GPS says one thing, the package’s actual location might be different.

According to research from McKinsey, GPS-enabled fleet tracking has improved delivery efficiency by 15-25% for logistics companies that implement it well. Fewer wrong turns, better route optimization, and accurate ETAs all flow from knowing where vehicles are.

Mobile Apps and Courier Check-ins

Modern courier systems use smartphone apps that drivers interact with throughout their routes. When a package is delivered, the courier marks it delivered in the app, often taking a photo as proof of delivery and collecting GPS coordinates automatically.

These apps provide much more granular data than just vehicle tracking. They capture arrival times, successful deliveries, delivery exceptions (customer not home, wrong address, etc.), and even customer signatures when required.

I’ve talked to several couriers about these systems. Most appreciate the structure and accountability, though some find the constant app interactions disruptive when they’re trying to move quickly through deliveries.

The apps also enable customer communication. Some logistics companies allow customers to message their courier directly through the tracking interface, asking for specific delivery instructions or confirming timing. This direct connection reduces failed deliveries.

Data Integration and APIs

Behind the scenes, tracking systems involve complex data integration. When you order something on Tokopedia and it ships via JNE, Tokopedia’s system talks to JNE’s system through an API (application programming interface), requesting tracking updates periodically.

These APIs allow e-commerce platforms to display shipping information without manual data entry. But they also introduce potential points of failure. If the API connection breaks, tracking information on the e-commerce platform might lag behind what the courier’s own system shows.

Team400, an AI consultancy I’ve encountered, works with logistics companies on data integration challenges like these. Getting disparate systems to communicate reliably isn’t trivial, especially when each company uses different database structures and update frequencies.

Predictive ETAs

The estimated delivery dates you see aren’t just guesses. They’re calculated based on historical performance data, current package location, destination, known route transit times, and sometimes even traffic data or weather forecasts.

Machine learning models have made these predictions significantly more accurate. Instead of generic “2-3 business days,” systems can now predict delivery windows within a few hours. Algorithms learn that packages from warehouse A to postal code B typically take X hours, accounting for variables like day of week and seasonal demand patterns.

When I order something in Jakarta for delivery within the city, J&T’s system often predicts delivery to within a 2-3 hour window. That’s impressive accuracy that wouldn’t be possible without sophisticated data analysis.

International Tracking Complexities

International package tracking involves multiple carriers handing off to each other. Your package might go from a US seller to USPS to Indonesian customs to Pos Indonesia to final delivery. Each organization has its own tracking system.

Universal tracking numbers help, but information gaps are common during handoffs. The package might show as “departed origin country” for days before updating to “arrived in Indonesia” because customs processing isn’t always tracked in real-time.

Customs adds another layer of unpredictability. Packages can sit in customs for days or weeks depending on inspection queues, paperwork issues, or random selection for detailed inspection. Tracking might just say “customs clearance in progress” without much detail.

What Goes Wrong and Why

Tracking failures usually stem from a few common issues. Missed scans create information gaps—the package is moving, but the system doesn’t know. System delays mean scans are captured but don’t upload to the central database immediately due to internet connectivity issues or server problems.

Package misrouting can make tracking information confusing. If your package accidentally goes to the wrong city, tracking might show seemingly backward progress as it gets rerouted correctly.

Barcode damage is another problem. If the barcode gets dirty, torn, or water-damaged, scanners can’t read it. The package might then be processed manually without tracking updates until someone replaces the label.

Very occasionally, database errors cause tracking numbers to show completely wrong information or fail to display at all. These are usually resolved quickly once reported, but they’re confusing when they occur.

Privacy Considerations

Real-time tracking requires collecting substantial data about package movements, delivery addresses, and courier locations. This raises privacy questions that logistics companies are increasingly thinking about.

Most companies now anonymize courier GPS data after delivery completion and limit how much location detail is shared publicly. You can see that your package is “in your area” without necessarily seeing the exact GPS coordinates of the delivery vehicle.

Customer delivery addresses are obviously sensitive data. Proper database security is essential to prevent this information from being leaked or misused.

The Future of Tracking

Expect tracking to get more detailed and accurate. Some companies are experimenting with package-level GPS trackers rather than just vehicle tracking. These small, reusable devices attach to high-value packages and report location independently.

Computer vision and AI are enabling automated package identification without requiring manual scans. Cameras at sorting facilities can recognize packages and update tracking automatically as they move through the system.

Customer communication will likely become more proactive. Instead of you checking tracking, systems will message you with updates: “Your package is 15 minutes away” or “Delivery attempted, but you weren’t home—would you like to reschedule?”

Integration with smart home devices is starting to happen too. Imagine your doorbell camera automatically notifying the tracking system when a package is delivered, instantly updating status without courier input.

Real-time package tracking has gone from impossible to unreliable to mostly dependable in just about 15 years. The technology will keep improving, making delivery more transparent and predictable for everyone involved.