How can I use Visual Quality Inspection to stop defective products during Quality Control?
The traditional way of visual inspection comprises of direct, manual examination using the naked eye, sometimes with simple tools like a magnifying glass or flashlight, to look for surface defects, wear, and inconsistencies. To reduce the fatigue and variability associated with human inspection, Luxio-AI system addresses this by using cameras and software to continuously monitor products for production lines. It utilizes artificial intelligence and computer vision to inspect 100% of the products in real time, automatically identifying defects like scratches, cracks, or color mismatches.
Can your AI models be trained for specific or unique products?
Yes, absolutely. Creating custom models for unique products is often a “black box” process with limited client input, leading to models that miss subtle defect nuances. The Luxio-AI system is designed to be trained on specific products, where reference images of both good and defective items are gathered to create a custom AI model tailored precisely to required quality standards. Once this foundation is established, the PANDORA AI platform is available to provide full autonomy for internal teams. This code-free interface allows operators to independently Label, Train, and Deploy model updates as production evolves.
How can we automate the rejection process to avoid manual sorting?
The manual sorting rejection process involves human operators identifying items that fail to meet specific criteria, removing them from the main flow, and placing them into a designated “reject” stream for further handling, analysis, or disposal. To create a fully closed loop quality control system, the PANDORA AI communicates with hardware by integrating directly with PLC(Programmable Logic Controller). When a defect is spotted, the system triggers sorting mechanisms like robotic arms, air jets, or pusher rejectors to remove the item from the conveyor belt without human intervention.
How does this technology handle subtle or organic defects that standard cameras usually miss?
Traditional Rule Based Machine Vision only looks for fixed measurements like width or height, which often fails with variable products. In contrast, Deep Learning models that are used in Luxio-AI learn what a good product looks like from examples. This allows the system to spot subtle or variable defects, such as organic textures in food, unpredictable scratches on metal, or complex patterns, much more effectively than older systems.
What is the difference between IoT, IIoT, and AIoT in your solutions?
These terms describe the environment and intelligence level of the devices:
IoT vs IIoT: While IoT is general, IIoT (Industrial Internet of Things) refers to ruggedized sensors designed specifically to withstand harsh factory conditions like dust, heat, and vibration.
AIoT (Artificial Intelligence of Things): This fuses AI with the Industrial Internet which allows devices like Luxio-AI to not just collect data but make autonomous decisions on the edge, such as rejecting a part, rather than just sending data to a server.
How do I get real time production counts without manual tallying?
Typically, to get real time production counts, a sensor is installed on the line that triggers a digital count every time a product passes. To streamline this process, Pandora Komis is installed directly on the conveyor belt and pushes data instantly to Pandora ERP, automatically updating the Product Receive Note.
How do I track key metrics like OEE?
Calculating OEE (Overall Equipment Effectiveness) requires continuously gathering data on availability, performance speed, and quality. The manual data collection typically involves operators using tick sheets, paper logs, or shared spreadsheets to record start/end times of stops, total counts, and rejected counts. However, Pandora platform automates this by centralizing data from devices like Pandora Komis (counters) and Pandora Revma (power monitors). This process provides live calculation of vital metrics, which aids in the identification of bottlenecks.
How can I predict machine failures before they stop production?
Traditionally, a machine couldn’t be identified that it was broken until it actually stopped production, which caused significant downtime. To solve this issue, Pandora Rival and Pandora Revma uses predictive maintenance that works by detecting minute physical changes, like vibrations or current spikes, that signal an upcoming breakdown. By tracking data like Vibration Analysis and Motor Current Signatures, the system detects tiny spikes, allowing maintenance to be scheduled before a failure occurs.
How can we eliminate data entry errors at the receiving bay?
The manual data entry process involves a human operator transcribing data from a source document into a computer system, which is prone to human error. To eliminate the high error rates from this process, it can be replaced with Pandora Varos (Wireless Smart Weighing) which uses wireless integration where the weighing scale speaks directly to the software. When raw materials are weighed, the data flows directly into the Material Receive Note in Pandora ERP, ensuring inventory levels are 100% accurate from the moment goods arrive.
How can we digitize our factory if we have older legacy machines?
Previously, adopting new technology often required the expensive purchase of entirely new machinery. However, Hexa IoT specialize in Legacy Machine Integration, connecting to existing PLCs or Human Machine Interfaces via standard protocols like Modbus or OPC UA. Hence, purchase of new machinery is not required as modern sensors can be retrofitted onto older equipment. This allows an upgrade to the factory without capital expense of replacing the current equipment.
How can I monitor soil health remotely to improve crop yield?
The manual process of monitoring soil health includes visual evaluation of soil structure, performing a simple infiltration test, and checking the soil’s smell and color. To remove the need for manual testing, Pandora Georgia placed connected sensors in the ground to transmit data to track critical metrics including Soil NPK, pH Levels, Temperature, and Moisture. This live data allows you to practice Precision Agriculture, applying fertilizer and water exactly where needed.
How can we monitor water quality in real time for aquaculture?
Effective aquaculture requires constant vigilance over water metrics to prevent stock loss. Manual water quality monitoring involves collecting representative samples, performing on-site tests, and sending samples to a lab for analysis that is time consuming. However, Pandora Thalassa measures Dissolved Oxygen, Salinity, Ammonia, and pH in real time. If oxygen levels drop, the system alerts immediately so that aerators can be activated instantly.
How do IoT systems function in remote areas without power lines?
In the past, automated monitoring was often impossible in remote areas because the lack of power infrastructure prevented equipment from functioning. For large plantations or remote ponds, systems must be self-sufficient. Hence, Pandora Georgia and Pandora Thalassa are fully Solar Powered with battery backups. They utilize long range wireless networks, such as LoRaWAN or 4G LTE, to transmit data, ensuring connectivity even where running power cables is impossible.
What makes HEXA IoT different from other IoT providers in Malaysia?
Many providers are hardware resellers, whereas we are Industrial Data Specialists and solution developers. We focus on providing practical, affordable solutions for SMEs. We build our own software (Pandora), integrate it with the best hardware, and customize it to solve your specific real world problems.
How do these solutions generate ROI for SMEs?
High ROI is achieved by focusing on modular systems that scale. Unlike large multinational integrators, we design specifically for SMEs, allowing you to start with a single sensor or camera and scale up. The system pays for itself through labor savings, reduced waste, and increased efficiency.
What is the process for starting a project with HEXA IoT?
We begin with a detailed discovery process. Our team works with you to define your current challenges, existing infrastructure, and the specific metrics you need to track. This ensures we propose a solution that is perfectly tailored to your needs.
How does the platform integrate with our existing software?
To maintain a clear view of your operations, it is crucial that new data flows into your current systems. The HEXA IoT Pandora Platform is built on an open architecture using RESTful APIs and standard industrial protocols. This allows us to push data from our sensors directly into your existing ERP, MES, or SCADA systems seamlessly.
