Background

Inside the production floor, a different kind of accuracy problem was present:

• Workers were manually identifying which type of material was in each bucket before weighing — fast, but unreliable.
• Misclassification meant production data for scrap materials (Kiriko and NG Wheel) couldn’t be fully trusted.
• There was no reliable way to trace which bucket held what material at any given moment.
• Without accurate classification data, optimizing the reprocessing of scrap material was largely guesswork.

Solution

I built a multi-hardware integration system that combined RFID tagging with industrial scale readings — removing the human identification step entirely. Each bucket would carry its own identity via an RFID tag, and the scale would capture the weight automatically through PLC integration. The system would then match the two data points together without any operator input.

Implementation

The RFID layer was implemented using the Zebra RFID SDK (Symbol.RFID3) in a VB.NET desktop application. Each physical bucket was registered in the system with a unique RFID tag linked to its material classification. When a bucket came in for weighing, the reader identified it automatically.

The industrial scale was connected via PLC, which performed a Direct Database Write the moment a weighing event occurred. A data correlation engine on the VB.NET side then paired the bucket’s RFID-derived identity with the PLC-derived weight, producing a complete, accurate record for every transaction.

A Laravel-based web dashboard was built on top of this to give management real-time visibility into production volumes — total Kiriko and NG Wheel material processed, with full traceability per transaction.

Impact

• Material misclassification dropped to zero — every database entry was backed by a hardware-verified RFID read and a direct PLC weight capture.
• Production teams gained instant visibility into scrap volumes, enabling more efficient reprocessing planning.
• Every transaction carried a complete digital trail — timestamp, material type, weight, and container ID — making audits straightforward.
• The combination of RFID and PLC data removed the last manual touchpoint in the classification and weighing workflow.

Background

At a manufacturing facility, every truck entering and leaving the site passed over a weighbridge — and every measurement was being recorded by hand:

• Operators wrote down weight readings manually, creating an obvious opportunity for recording errors or manipulation.
• Records existed only on paper, with no reliable audit trail for external inspectors.
• When the customs authority (Bea Cukai) required material reporting, the company had no digitally verifiable data to present.
• The administrative process around weighing was slow, paper-heavy, and entirely dependent on human accuracy.

Solution

Rather than building the system from scratch, I took over the ongoing development and maintenance of this Laravel-based web platform. Acting as the central control and monitoring hub, the system connected directly to the weighbridge hardware through a PLC (Programmable Logic Controller). I took full responsibility for ensuring the platform reliably captured weight data automatically the moment a truck was on the scale — keeping human involvement out of the measurement and preventing any post-weighing alterations.

Implementation

• Hardware Integration & Security: I maintained and optimized the hardware-to-web bridge that received live data from the PLC sensor and wrote it directly to the database, completely bypassing manual input. I also managed the system’s anti-tamper validation layer, ensuring that only data originating strictly from the sensor could be recorded, effectively blocking any attempts to modify weight values through the user interface.
• Audit-Ready Data & Reporting: I oversaw the database architecture to ensure it remained strictly audit-ready, where every transaction carried a clear, traceable record for verification by external authorities. Furthermore, I managed and refined the automated material recapitulation module, which classified incoming and outgoing materials by type, vendor, and tonnage, providing the management team with a live, accurate view of the logistics flow without needing to dig through raw records.

Impact

• Human input was fully removed from the measurement process, eliminating the risk of recording errors and manipulation.
• Customs inspections became significantly faster — every data point was already centralized, timestamped, and accessible through the web platform.
• Management gained instant visibility into material volumes entering the facility, enabling more precise production planning.
• The accuracy of financial data tied to material purchases improved substantially as a direct result of automated weight capture.

Background

As the company’s software products grew and reached more clients, a quiet but serious problem began to surface:

• Every new client activation had to be handled manually — sending license keys, tracking expiry dates, pushing updates one by one.
• There was no central system to oversee which devices were running which version.
• No reliable way to cut off access when a license expired or a device was reported lost.
• The process was fragile, error-prone, and simply didn’t scale.

Solution

The answer was a Centralized License Management Hub — a system that could govern the entire software lifecycle from one place. Rather than patching the existing manual process, I designed it from scratch to be modular and agnostic, meaning it could be plugged into any desktop application the company built, not just one product.

Implementation

The system was built on a Laravel backend exposing RESTful APIs, integrated with VB.NET desktop clients on the user side. It supports two license tiers — Full Access and Customizable Trial — with configurable validity periods that can be adjusted on the fly without redeployment.

On the security side, every license validation happens in real time. Each device is fingerprinted at the hardware level, so a license can’t simply be copied to another machine. If a device is flagged — whether for policy violation or being reported missing — it can be remotely blacklisted and locked out instantly.

Distribution was handled through an automated update engine written in Bash Scripting, which pushed the latest version to all registered devices simultaneously. A monitoring dashboard gave the team a live view of which client owned which devices, what license type they were running, and when it was set to expire.

Impact

• Manual intervention in activation and maintenance dropped by 80%.
• Version consistency improved to 100% across the client ecosystem — no more compatibility issues from outdated builds on forgotten machines.
• The modular architecture allowed the same system to be extended to other company products without rebuilding anything from scratch.
• Remote blacklisting gave the team an immediate response mechanism for lost or misused devices.