RTD sensors as Industrial IOT sensor -Vision AI OCR applications

Vision AI on digital Display in Sembang AIoT live Channel

Vision AI OCR on water meter indicator

The extended presentation of “Sembang AIoT” commenced with a focus on Vision AI OCR. In this context, digital meters served as a primary data source for the AI OCR engine, facilitating the extraction of text from various meter types and indicators.

Furthermore, the discussion revolved around the concept of non-intrusive data extraction, prompting a recollection of our initial non-intrusive data extraction kits, which were showcased at an IoT summit event in Penang hosted by Intel Malaysia. During this demonstration, we exhibited non-intrusive data extraction using an industrial ebox IPC system to intercept VGA output from simulated machine data. Subsequently, video content was captured and extracted using an OCR engine running on the ebox IPC, as depicted in the accompanying image. Additional visual material can be found on our parent company Axiomtek Malaysia’s  Facebook page by searching for “non-intrusive.”

The live demonstration proceeded to encompass a new array of meters and indicators, referred to as “Digital Data Extraction within Digital.” This phase involved scanning all digital indicators with a camera and utilizing the Intel Openvino AI engine, which was operational on the Axiomtek AI Edge computer, to perform AI OCR on diverse digital displays for data extraction.

For further insights and interactive explanations, please visit our live channel and explore additional examples.

The demonstration of the Vision OCR AI on the Temperature controllers. On the right, the pressure transmission indicator is used to let the Axiomtek AI edge perform the AI OCR inferencing.

see more interactive explanations in the live channel. 

RTD sensors IIoT applications -2 wires, 3 wires and 4 wires explained

The demonstration included a dedicated discussion on industrial RTD (Resistance Temperature Detector) temperature sensors, highlighting their critical relevance in practical industry applications. Our aim was to provide in-depth insights into the usage of 2-wire, 3-wire, and 4-wire RTD sensors, elucidating the distinctions, advantages, disadvantages, and the considerations involved in selecting the most suitable RTD sensor for a specific application.

Temperature stands as a pivotal parameter in numerous industrial processes. Consequently, it assumes paramount importance for Industrial Internet of Things (IIoT) applications, necessitating an accurate understanding and appropriate utilization within the realm of measurement.

To gain a comprehensive understanding of these topics, we encourage you to explore our interactive live session, commencing from the 27th minute onwards. You can locate this session by conducting a search on YouTube using the query “YouTube #21 Sembang AIoT.” Should you encounter any difficulty in locating the content, please do not hesitate to inform us, and we will provide further assistance.

2-wire, 3-wire, and 4-wire RTD (Resistance Temperature Detector) sensors are commonly used for measuring temperature in various industrial and scientific applications. The main differences between them lie in their wiring configurations and their impact on measurement accuracy and compensation for lead wire resistance. Here are the key differences and their typical applications:

2-Wire RTD Sensors:

Wiring Configuration: A 2-wire RTD sensor has two wires: one for the RTD element and one for the lead wire.
Advantages:
Simplicity of wiring.
Cost-effective.
Disadvantages:
Less accurate compared to 3-wire and 4-wire configurations due to the inability to compensate for lead wire resistance.
Applications: Used in applications where high accuracy is not critical, and cost and simplicity are primary concerns. Common in basic temperature measurement tasks.
3-Wire RTD Sensors:

Wiring Configuration: A 3-wire RTD sensor has three wires: two for the RTD element and one for the lead wire. The two wires for the RTD element are connected in a series.
Advantages:
Provides compensation for lead wire resistance, improving accuracy.
Offers good accuracy while remaining cost-effective.
Disadvantages:
Not as accurate as 4-wire configurations but suitable for many applications.
Applications: Used in applications where moderate accuracy is required, and compensation for lead wire resistance is necessary. Common in industrial temperature monitoring and control.
4-Wire RTD Sensors:

Wiring Configuration: A 4-wire RTD sensor has four wires: two for the RTD element and two for the lead wire. The two sets of wires are connected in parallel.
Advantages:
Highest accuracy among the three configurations.
Accurate compensation for lead wire resistance.
Disadvantages:
More complex wiring and potentially higher cost.
Applications: Ideal for applications demanding high accuracy and precision, such as laboratory and scientific measurements, calibration standards, and critical industrial processes. 

In summary, the choice between 2-wire, 3-wire, and 4-wire RTD sensors depends on the level of accuracy required, the budget constraints, and the specific application. While 2-wire RTDs are the simplest and most cost-effective, they are less accurate. 3-wire RTDs strike a balance between cost and accuracy, making them suitable for many industrial applications. 4-wire RTDs provide the highest accuracy but come at a higher cost and complexity, making them suitable for precision measurements and calibration standards.

AIoT Edge Connect IIoT cloud platform featuring:-

  1. Data Visualization on the dashboard
  2. Historian Data logging
  3. Alert management ( Whatsapp, SMS , Telegram and Email )

watch us live at the link blow: 

 

 

 

Practical AI OCR and IIoT Web API hosting

Unlocking the Future of Manufacturing with AI and IoT: Axiomtek’s Latest Breakthrough

Welcome back to Sembang AIoT, your go-to source for all things AI and IoT! In our 20th edition, we’re about to embark on an exciting journey where technology meets innovation.  Axiomtek Malaysia with AIoTmission has just unveiled a game-changing revelation, one that’s set to redefine the landscape of manufacturing and beyond.

We’ve got more in store for you, and this time, it’s all about the Internet of Things (IoT) sensors, specifically tailored for Smart Agriculture. We’re talking about a lineup of sensors that includes everything from pH and salinity sensors to soil NPK sensors, temperature gauges, and humidity detectors. What makes these sensors even more remarkable is that they can seamlessly communicate and relay data to the AIoT Edge Connect Cloud, local OPCUA servers, local Web APIs, or even web servers, making data accessibility easier than ever before.

But the real magic happens when you witness our Axiomtek IIoT Edge Gateway in action. This powerful device not only collects data from the RTD Temperature module and various sensors but also logs this valuable information at the edge gateway. The data isn’t just stashed away; it’s sent to a Web API or Rest API server, residing right on the Edge Gateway.

Now, imagine this scenario: you’re miles away from the production floor, perhaps in another corner of the factory or even off-site. You need real-time data, and you need it now. That’s where our Adisra IIoT SCADA software comes into play. It’s designed to seamlessly connect to the Web API server, enabling you to access live data online. Plus, you have the power to schedule data retrieval as frequently as needed. Need updates every second? No problem, we’ve got you covered.

At the core of this innovation is the idea that data is not just information but a valuable resource that can transform the way we manufacture, manage, and optimize processes. With Axiomtek’s cutting-edge technologies, you’re not just embracing the future; you’re shaping it.

 

Axiomtek IIoT edge gateway is hosting the data polled, in this case, 2 x RTD sensors data from the RTD module were obtained. Local display on the temperature was shown on the gateway as an option nevertheless the IIoT edge gateway can be a stand alone headless if you want it to be.

The right hand picture showed on another IPC running Adisra SCADA software with Web API client driver where the http get can be done in a quite straightforward manner. 

In this demo, we only apply user name and password to connect to the Web API server although it supports token + User name+ Password as well. 

The latter part of the presentation commenced by delving into the AI OCR model, offering insights into its background and development. We are leveraging a highly efficient model capable of optical character recognition (OCR) through computer vision, primarily focusing on two key objects adorned with printed characters. These objects include the “Sirim” sticker found on a 3-PIN power plug and the printed content on the Axiomtek carton box.

Picture this: Real Vision OCR (Optical Character Recognition) powered by the formidable OpenVINO AI engine from Intel, all seamlessly integrated into the groundbreaking Axiomtek AI Edge Platform..

Sirim STandard approval sticker on Vision AI camera. Character read and recorded

AI OCR (Artificial Intelligence Optical Character Recognition) has several applications in the manufacturing industry, where it can improve efficiency, accuracy, and automation. Here are some key applications of AI OCR in manufacturing:

1. Quality Control and Inspection:

– AI OCR can be used to inspect product labels, barcodes, serial numbers, and other critical information to ensure they meet quality standards.

– It can identify defects or discrepancies in printed labels or packaging, helping to reduce product recalls and maintain quality control.

2. Inventory Management:

– AI OCR can automate the process of tracking and managing inventory by recognizing and updating product information from labels and packaging.

– It can help prevent overstocking or understocking of items and improve supply chain management.

3. Document Management:

– Manufacturers deal with various documents, such as invoices, purchase orders, and shipping labels. AI OCR can automate the extraction and digitization of data from these documents.

– This improves accuracy and reduces manual data entry, saving time and reducing errors.

4. Equipment Maintenance:

– AI OCR can be used to read and interpret data from equipment sensors and gauges, allowing for predictive maintenance.

– By analyzing this data, AI systems can predict when machinery needs maintenance or repairs, reducing downtime and maintenance costs.

5. Regulatory Compliance:

– Manufacturers often need to comply with various regulatory standards and certifications. AI OCR can assist in the verification and validation of documents and labels to ensure compliance.

– It can flag discrepancies or missing information that could lead to compliance issues.

6. Traceability and Serialization:

– In industries like pharmaceuticals and food production, traceability is crucial. AI OCR can help in tracking and tracing products throughout the manufacturing process.

– It can read and verify serial numbers, lot codes, and other identifiers to ensure product authenticity and traceability.

7. Supplier and Vendor Management:

– AI OCR can automate the verification of supplier and vendor invoices, ensuring that they match the agreed-upon terms and quantities.

– It can also help in tracking payments and managing accounts payable more efficiently.

8. Maintenance and Repair Manuals:

– AI OCR can convert printed maintenance and repair manuals into digital formats, making it easier for technicians to access and search for information.

– It can also be used to extract relevant data from these manuals for troubleshooting and repair purposes.

9. Energy Management:

– Manufacturers can use AI OCR to monitor energy consumption by reading and analyzing utility bills and meter data.

– This information can help in optimizing energy usage and reducing operational costs.

10. Labeling and Packaging:

– AI OCR can assist in automated labeling and packaging processes, ensuring that labels are applied accurately and products are packaged correctly.

– It can also check for alignment, print quality, and the presence of required information on product packaging.

In summary, AI OCR plays a vital role in streamlining various aspects of manufacturing, from quality control and inventory management to compliance and documentation. It enhances efficiency, reduces errors, and contributes to overall cost savings in the manufacturing industry.


Watching us  live or future live session by subscribing to our youtube channel below:-

https://youtube.com/live/FXHJZn2J3Fs






Hosting IoT sensor data on OPCUA Server and presenting AI data

Presenting vision AI data and Hosting IIoT sensor data with OPCUA SERVER

Welcome to the 19th episode of our thrilling AIoT Live Series! We’re not just discussing AIoT – we’re bringing it to life with real-world demonstrations and insights. Witness firsthand how AIoT can amplify your projects.
 
In our latest session, we delved deep into Vision AI techniques. We showcased how AI extracts data from images and the captivating ways in which this data is interpreted and presented.
 
Wondering how the AI engine displays results? Here’s a sneak peek:
 
  1. 🖼 Annotations directly on images/videos.
  2. 📝 Textual descriptions or captions.
  3. 🎧 Audible outputs.
 
These are the basics of AI data representation. As a real-world example, consider a moving vehicle on a road. Curious about identifying the object, its size, or tracking its identity? The AI engine has got it all sorted, once the right model is in place.
 
Whether it’s the nimble AI edge system from Axiomtek or a robust AI server engine in a control room, the choice of computing engine can cater to various needs.
After the AI completes its inferencing, it visually communicates using a ‘bounding box’. Think of it as AI’s way of saying, “Look here!” In our example, the spotlight is on a truck – that’s our identified object.
 
Want more insights? 🎥 Dive into our discussions on our YouTube channel:

Live demo on OPCUA SERVER to host the IIOT sensor data

Diving Deeper into IIoT: The Key Instruments & Protocols 🌐

 
In the second segment of our live session, we ventured into the world of process controllers, transmitters, and analyzers pivotal to the industry. These IIoT data points play a central role in strategic decisions for efficient process control plant operations.
 
The communication backbone of these systems often rests on two primary protocols:
 
  1. Modbus (Previously Modicon Bus, now under Schneider Electric)
    • Platforms for Modbus communication include:
      • TCP/IP over Ethernet.
      • Asynchronous serial communication encompassing diverse standards and technologies like EIA/TIA-232-E, EIA-422, EIA/TIA-485-A, and even extending to fiber and radio frequency.
  2. HART (Highway Addressable Remote Transducer)
    • HART stands out as a hybrid analog+digital open protocol for industrial automation. It shines with its capability to operate over the traditional 4–20 mA analog loops, leveraging the existing wire infrastructure of analog-only systems. With its adaptability, HART finds its place from modest automation tasks to intricate industrial processes.
 
When it comes to integration, Modbus typically offers a smoother transition to IIoT gateways. However, with HART, a bridge is often required – a transceiver or converter that transitions from HART to Modbus, ensuring the captured HART data remains organized and intact.
 
Stay connected for more insights into the ever-evolving realm of IIoT! 🛠
OPC UA Server Demo: A Deep Dive into Data Configuration & Sharing 🌐

We kicked off our demonstration showcasing the setup of an RTD sensor module, integrating it seamlessly with a local embedded SCADA station. Acting as the data nexus, this gateway gathers all data via the Modbus interface. Subsequently, the data finds its way to the OPC UA server, where it’s hosted and configured to be accessible to other OPC UA clients.

For this demonstration, we utilized the ‘Unified Automation OPC UA client’, a renowned third-party client, to connect to the OPC UA server hosted on the Axiomtek Edge gateway.

Eager to see the demonstration in action? Dive deep into our full showcase on our YouTube channel. Join us for an immersive experience! 🎥🚀


AIoT in workspace safety and sensor integration

The “Sembang AIoT 18” sharing session began with a deep dive into AI model-building essentials. Kien Leong emphasized the significance of Dataset augmentation in the realm of computer vision AI.
 
Reflecting on the Axiomtek AIS (AI suite) utilized at Plug Fest II, this suite was an integral part of the training initiative by the Malaysia Productivity Centre and EEPN. A key feature of the AIS was its Dataset augmentation capability. This enabled the raw dataset to undergo an augmentation process, essentially scrambling and transforming the data to better train the AI model, enabling it to deeply learn various appearances and combinations.
 
Additionally, we showcased a brief demonstration of computer vision AI focusing on workspace safety. This AI was adept at monitoring compliance factors like the wearing of helmets, vests, and more.
 
The AI system ran on one of Axiomtek’s AI edge/IPC platforms.
 
A major part of our discussion revolved around the challenges faced, emphasizing that the dataset’s quality from the site directly affects the accuracy of AI model decisions. For more insights from this session, watch the recorded live session.
IIOT pressure Sensors to Cloud

The pressure, an important  process parameter that require accurate measurement must not be forgotten in one of the IIoT sensors where it is widely used in the industries. 

We highlighted two types of commonly used sensor namely, Strain gauge sensors and Piezoelectric sensors. These sensors are used in many instrument and apparatus.

Piezoelectric Sensor:

Principle: Piezoelectricity is the electric charge that accumulates in certain solid materials (like crystals, certain ceramics, and biological matter like bone) in response to applied mechanical stress. The word “piezoelectricity” means electricity resulting from pressure.

Function: When pressure is applied to a piezoelectric material, it generates a voltage proportional to that pressure. This voltage can then be measured and used to determine the amount of pressure applied.

Applications: Due to their ability to respond to fast changes in pressure, piezoelectric sensors are often used in dynamic applications. They are found in many applications including:

Measuring dynamic pressure changes in gases and liquids.

Detecting the force of impact in automotive airbag systems.

In musical pickups, especially for acoustic instruments.

In some microphones to detect sound waves.

Strain Gauge Pressure Sensors:

Principle: A strain gauge consists of a conductive pattern (often made of a metallic foil) that is affixed to a flexible backing. When this pattern is stretched or compressed, its electrical resistance changes.

Function: When pressure is exerted on a diaphragm or membrane, it deforms (or strains). A strain gauge attached to this diaphragm will also deform, leading to a change in its resistance. This change in resistance is proportional to the pressure exerted on the diaphragm. By measuring this change in resistance, the applied pressure can be determined.

Applications: Strain gauge pressure sensors are widely used due to their reliability and stability. They are found in various applications including:

Industrial process monitoring and control.

Aerospace and automotive sensors.

Medical devices for monitoring blood or fluid pressures.

Load cells for weight measurement.

Both types of sensors come with their own sets of advantages and limitations. While piezoelectric sensors excel at measuring dynamic pressures over short durations, strain gauge sensors are often preferred for stable, continuous pressure measurements.

Watch in detail sharing on the Aiotmission youtube channel with the link below: 

 

 

AIOT sensor to Cloud and data analytic

AIoT sensors to Cloud & Data Analytic

In manufacturing, numerous processes necessitate controlled heating, underscoring the paramount importance of accurate temperature measurement. Processes such as metal forging, welding, boiling, pasteurization, drying, and many others demand meticulous temperature regulation to guarantee superior product quality.


What are the principal sensors utilized for temperature measurement? Furthermore, which IIoT devices facilitate the digitization of these readings for enhanced monitoring and control? Which tools offer optimal visualization coupled with robust data analytics at the cloud level? These critical topics were the focal points of our recent “Sembang AioT session.” This discussion aimed to shed light on essential temperature-measuring instruments and advanced IIoT devices, underlining their significance in ensuring precise temperature control and, consequently, contributing to the production of high-caliber products.

There are three types of temperature sensors that are widely used in industries. 

1. Thermocouples (TCs):

Working Principle: Based on the Seebeck effect, where a voltage is produced when two different metals are joined together and the junction is heated.

Application: Used in various industrial applications due to their wide temperature range, ruggedness, and accuracy.

2. Resistance Temperature Detectors (RTDs):

Working Principle: RTDs work on the principle that the electrical resistance of a material (typically platinum) changes with temperature.

Application: Used in precision temperature measurement applications as they offer high accuracy and stability over a narrow temperature range.

3. Thermistors:

Working Principle: Thermistors are temperature-sensitive resistors whose resistance changes significantly with temperature. There are two types – NTC (negative temperature coefficient) thermistors and PTC (positive temperature coefficient) thermistors.

Application: Used for accurate temperature measurements in a limited temperature range.

Their applications are as follows:-

Thermocouples are suitable for measuring high temperatures and are used in various industrial applications, such as furnaces and processing plants. Wide temperature range -200 C to 1800 degree C.

RTDs offer high accuracy and stability for lower temperature measurements and are widely used in scientific and industrial settings. The range of the temperature is from -250 C to 1000 Degree C.

 

Thermistors are typically used in low to moderate temperature measurements, including climate control and medical devices. The temperature ranges from -55 C to 114 degree C

After having the right temperature sensors elements, next is to have a suitable IO module that allows you to do measurement. The picture beside this shows some Temperature module (Thermocouple input & RTD sernsor input module)  that supports Modbus RTU or Modbus TCP that provide a interface on serial or Lan port to be connected to the Axiomtek IIoT edge gateway or the AEC310 4G router gateway .

There are several way of how to treat the data collected. One of them is to allow the IIoT edge gateway or another 4G gateway to acquire from it, like what it is showed in picture at the right hand side 

As for Cloud wise, AIoT Edge Connect cloud platform that is sitting on the Microsoft Azure can be used to receive all data collected and can then be shown on the Cloud Dashboard.

In this live session, Kien Leong delved into the requirements and standards for AI datasets used in training and deep learning.

The critical role of datasets was discussed, emphasizing not only their utility for AI training but also their essential function in testing and evaluating the accuracy of AI models. A suggested dataset distribution was presented, advocating for a 70-20-10 proportion to enhance practical AI deployment. This entails allocating 70% of the dataset for training, 20% for validation, and the remaining 10% for testing to ascertain model accuracy, a configuration applicable to most AI applications. Further discussion was centered around the different classes of datasets.

For a comprehensive insight into these crucial topics, you are encouraged to watch the live session. Ensure to subscribe to the channel using the link below to stay abreast with forthcoming intriguing discussions on AIoT.

Please watch this at the link below:- 

https://youtube.com/live/ADSdC9QkT4s

 

 

Smart Nation Expo 2023 AIOT demonstration

In collaboration with Axiomtek Malaysia at the Smart Nation 2023 Expo, we presented live demos featuring the most recent advancements in Industrial Internet of Things (IIoT) and Artificial Intelligence (AI).
 
Our additional endeavors include investigating, studying, and exploring AI and IIoT sensors, which enable us to integrate AIoT into a variety of industrial applications

AIoTmission took this opportunity to create several case studies and actual applications in the industries. Starting with Vision AI on supervised and unsupervised Deep learning application in object detection and anomaly detection in quality control. It is then followed by the industrial IoT computing platforms such as IIoT edge gateway and edge HMI that established a data collection and communication layer of the OT ( Operation Technology) in the industrial IoT eco system.

In the Live sharing ” Jom! Lets Sembang AIoT” , We did a recap of all the demo kits that we showed in the expo. 

Mr. CC Lee, shared some of the sensors relating to measurement such as Ultrasonic level sensor and  Radar level sensors as in the process of digitization, right sensors is very important in the whole measurement that then provide a accurate dataset for machine learning or AI data analytics . 

Ultrasonics Level sensors Vs Radar Sensors

Some pros and cons of the sensors and the applications are briefly covered as below:-

Environmental Conditions: 

Ultrasonic transmitters should be mounted in a predictable environment. This is because dust, humidity, and other physical parameters may contaminate and affect the accuracy of the reflected signal. On the other hand, radar transmitters work well even in a dusty and overall harsh industrial environment.

 Pressure Limits: Ultrasonic transmitters are not intended for high or negative pressure limits. The device can bear a maximum working pressure of 3 bar. But, radar may work in maximum operating pressure over 4Mpa.

Radar has got better adaptation in higher pressure 

 Temperature Limits: Ultrasonic transmitters work well with temperatures over 80°C. Varying process temperatures or fluctuations may produce inappropriate readings. Guided wave radar level transmitter works well in temperature up to 315°C.

 Accuracy: Changes in density, acidity, and viscosity do not affect the accuracy of radar-level transmitters. Thus, radar transmitters are more precise than ultrasonic transmitters. For storage tank level measurement, high-accuracy radar transmitters are used.

Performance: The performance of the ultrasonic transmitter is based on the strength of the reflected sound wave, while radar performs well independent of the process conditions.

Applications: Ultrasonic transmitters are an excellent choice for solid as well as liquid level measurement. They are widely used for presence detection and object profiling. The car wash industry is one great example of using ultrasonic sensors to enhance efficiencies and improve processes. The most common applications of radar transmitters are radar transmitters are minerals and mining, oil and gas, asphalt blending tanks, alum and wax tanks, pharmaceutical, pulp and paper, and more.

ICPDAS Stack light module + simulated counter / performance input + Tower light indicators. 

connected wireless to Axiomtek Edge HMI 

non-intrusive data extraction via camera

Voltmeter Reading with  camera + KIWI300 embedded 2 cores intel celeron performing the Vision AI 

Pressure meter Reading with  camera + ICO100 iioT edge gateway 2 cores intel celeron performing the Vision AI 

vibration sensors + Power meters to AEC300 4G router gateway publishing to AioT edge connect Cloud 

To see more detail of the the Live Demo, Watch us live at our youtube Channel with the link below:-


https://youtube.com/live/TJLF1GIsFBY



AIoT smart manufacturing for SME

While "sembang AIoT" We celebrate Malaysia Day

Challenges with SME in the implementing Smart Manufacturing

This weekly live session, we shared some challenges faced by the SME in the IR4 journey and solutions to it followed by Vision AI application in a non intrusive analogue meter reading where in some cases, this solution is needed with least man power involvement and less error created by the human in the data collection.

Industry 4.0 (IR4) represents the Fourth Industrial Revolution, characterized by a range of new technologies, including the Internet of Things (IoT), artificial intelligence (AI), robotics, augmented reality, and more. While these technologies have the potential to transform manufacturing processes, SME (Small and Medium-sized Enterprises) manufacturers face specific challenges in their IR4 journey:

High Initial Investment: Deploying state-of-the-art technologies often requires significant financial outlay, which can be prohibitive for SMEs. This includes the costs of acquiring, installing, and maintaining new machinery and software.

Skills and Expertise: Implementing and maintaining IR4 technologies require specialized skills. SMEs may struggle to find or afford the right talent.

Cybersecurity Concerns: With increased connectivity and digitization comes the need for advanced cybersecurity measures. SMEs may lack the resources or expertise to effectively guard against sophisticated cyber threats.

ROI Uncertainty: The benefits of some Industry 4.0 technologies might not be immediately evident, making it difficult for SMEs to justify the investments based on potential returns.

Knowledge Gap: SMEs may lack awareness of the full range of technologies available, their potential benefits, and how best to implement them.

For SME manufacturers, the journey toward Industry 4.0 adoption is fraught with challenges, but the potential benefits in terms of efficiency, productivity, and competitiveness are considerable. With the right strategies, partnerships, and investments, these challenges can be addressed, paving the way for successful digital transformation.

 

you may find the in our previous OEE tracking live sharing, This is one of the demonstration that handle a full sets of OEE tracking in the production floor. 

Availability :  signaling from Tower Light

Performance:  the production count and cycle time 

Quality :  user interface to input the defects and type of defects .

Two main component in this Axiomtek IIoT edge HMI  and Stack Light module.

 

Non intrusive Analogue meter reading with Vision AI

non-intrusive data extraction via camera

The Live session ended with non intrusive meters reading. We showed the analogue Voltmeter reading with one of the camera with the AI engine running on KIWI310 ( a tiny embedded Intel based platform over USB), The data is visualized over the cloud dashboard.

We reserve the other pressure gauge Vision AI reading with the IIOT edge gateway to show case in the upcoming Smart Nation 2023 exhibition. You can find us at the exhibition in Hall 8 : booth 8043 at MITEC from 19th to 21st of Sept. 

Watch us Live in Youtube with the link below and appreciate your subscription to AioTmission youtube channel with the link blow.



https://youtube.com/live/6zbs3zPNALY

Progression from 1G to 5G Impacts AIoT Applications in Industries

Progression of 1G to 5G mobile network that impacts the AIOT implementation in the industry.

This week’s live discussion offers a unique conversation, or ‘sembang,’ exploring the transformative journey of mobile networks from 1G to 5G. We delved into how these advancements in mobile communications have profoundly impacted applications of AIoT across various industries. Additionally, we touched on Malaysia’s NEW INDUSTRIAL MASTER PLAN 2030 (NIMP2030).
 
Manufacturing continues to serve as a pivotal growth catalyst for Malaysia. With its significant contribution to the nation’s economy, NIMP 2030 has been designed to elevate industrial capabilities to new heights. The plan is centered around key missions and enablers.
 
Missions:
 Elevating Economic Complexity
  • Paving the Way for a Digitally Vibrant Nation
  • Striving for Net Zero Emissions
  • Ensuring Economic Security and Inclusivity
 Enablers:
 Mobilizing a Robust Financing Ecosystem
  • Encouraging Talent Development and Attraction
  • Streamlining the Investor Journey for Business Ease
  • Implementing a Nationwide Governance Framework
 At Sine AIoTmission, our focus lies in fostering innovation and technological development in the realms of Industrial Internet of Things (IIoT) and Artificial Intelligence (AI). Specifically, the mission of ‘teching up for a digitally vibrant nation’ and the enabler of ‘fostering talent development and attraction’ represent the facets where we are best positioned to contribute, especially in terms of digital transformation and talent cultivation.
The evolution of mobile networks from 1G to 5G has brought transformative changes to the Internet of Things (IoT) applications in various industries. Each generation brought with it improved speed, reliability, efficiency, and features that greatly influenced the capabilities and spread of IoT devices. Here’s a breakdown of how each generation impacted IoT applications:

  1. 1G (1st Generation)
    • Overview: 1G mobile networks introduced the world to cellular communication with analog voice.
    • IoT Impact: There wasn’t any direct impact on IoT since 1G was primarily voice-centric. However, it set the foundation for mobile communication which would be pivotal for IoT’s evolution.
  2. 2G (2nd Generation)
    • Overview: 2G introduced digital communication, leading to services like SMS and MMS. GSM, its most popular standard, had limited data service like GPRS and EDGE.
    • IoT Impact: 2G marked the beginning of Machine-to-Machine (M2M) communication, allowing devices to exchange data without human intervention. This facilitated the earliest versions of connected devices, such as fleet tracking systems.
  3. 3G (3rd Generation)
    • Overview: 3G brought faster data transfer rates, enabling mobile internet browsing, video calls, and mobile apps.
    • IoT Impact: The improved data speeds allowed for more advanced M2M communications. IoT devices could now transmit larger packets of data more efficiently, allowing for richer applications such as video surveillance, remote health monitoring, and more responsive remote monitoring and management systems.
  4. 4G (4th Generation) & LTE (Long-Term Evolution)
    • Overview: 4G offered even faster data speeds and lower latency, making real-time communication and HD video streaming on mobile devices possible.
    • IoT Impact: The data capabilities of 4G gave rise to a vast array of IoT applications. Smart homes, smart cities, connected cars, and industrial automation became feasible. Real-time data analytics, cloud integration, and more sophisticated remote applications saw significant growth.
  5. 5G (5th Generation)
    • Overview: 5G is characterized by ultra-high speeds, extremely low latency, increased bandwidth, and enhanced connection density.
    • IoT Impact: 5G has the potential to revolutionize IoT:
      • Ultra-reliable Low-latency Communication (URLLC): Crucial for applications like autonomous vehicles, industrial automation, and drone operations where split-second communication can be the difference between safe and unsafe operations.
      • Massive Machine Type Communication (mMTC): This allows for a massive number of devices to be connected in a small area, essential for smart cities or densely packed industrial environments.
      • Enhanced Mobile Broadband (eMBB): Provides high data rates for applications such as virtual reality (VR) or augmented reality (AR) in industrial training or maintenance.
      • IoT in Various Industries: With 5G’s capabilities, industries like healthcare can now use real-time remote robotic surgeries, agriculture can deploy precision farming techniques using sensors and drones, and smart cities can manage resources and traffic in real time.

The evolution from 1G to 5G has expanded the horizons of IoT applications in industries from basic remote monitoring to complex, real-time data-driven operations that can transform the way industries function. As we look to the future, continued advancements in mobile networks will further amplify the capabilities and integration of IoT in various sectors

some 2G modems with conventional RS232 interface and TcPip ( LAN) interface are shown in the session where this set of legacy at some points has brought many successful applications in for remote access in many industries.

The Live demo cover the followings : 

1. Onboard Modbus IO Devices to the 4G /5G gateway router

2. publish Data to the cloud 

3. Visualize data on the Dashboard

4. setting Alert -Whatsapp 

5. Data logging and Download

Ultimately, all the data collected will be used in the future AI Data analytic with any ready AI studio from Microsoft Azure, AWS, Google and etc. 

Watch this video on our Youtube Channel with the link below:-

 

Integrating Industrial IoT and AI for optimal fourth industry revolution performance

Integrating AI and IoT for optimal IR4.0 performance

This session of ” Sembang AIoT” talk more about the integrating IIoT and AI for optimal performance of industry 4.0 performance.

while IR4 principles lay the foundation for a connected, transparent, and autonomous industrial environment, it is the integration of AI and IoT that actualizes and optimizes these principles to achieve the full potential of Industry 4.0.
Industry 4.0 (IR4) is the current trend of automation and data exchange in manufacturing technologies, encompassing cyber-physical systems, the Internet of Things (IoT), cloud computing, and cognitive computing. Here are the four main design principles of Industry 4.0:

  • Interconnection 
  • Decentralization
  • Data Transparency
  • Technical Assistant

The four fundamental design principles provide a clear framework for Industry 4.0’s goals and future direction. Interconnection and Information Transparency are intrinsically linked to the Internet of Things (IoT), emphasizing the significance of connectivity. Data collected from sensors, machines, and controllers necessitates the use of IoT devices to maintain continuous connection and facilitate real-time data transfer to the cloud. This ensures consistent and transparent data throughout various processes.

https://axiomtek.my/product/ipc962-525-axiomtek-industrial-pc/This session is further enhanced with the properly set demonstration on the unsupervised Deep learning anomaly detection system. Axiomtek’s Fanless industrial PC is used in this deployment of demo. In order to speed up the inference process, one of the Entry level GPU is used to keep up with performance.

The unsupervised deep learning AI in this model do not require any intervention on the dataset that are used. Good dataset or samples are fed and with minimum work on the result that we need to achieve. We took one of the good example of Printed Circuit Board ( PCB) visual inspection process where the time required to check on the PCB takes about 45 seconds per board. But with Vision anomaly AI, It took only 2 seconds on the entire PCB with 4 to 6 cameras depending on the size of the PCB.

The picture on the left show the visual image of marked red AI spotted unwanted soldered joined and those images can be marked and stored as part of the defect listing.

Different zones of defects can be recorded for the next process of rework. 

 

Unsupervised deep learning is a subset of deep learning where models are trained without labeled data. The primary objective is to extract patterns or knowledge from the data without explicit guidance on what to look for. Here’s a breakdown of the concept:
 
Unsupervised Learning: In traditional machine learning, unsupervised learning refers to training algorithms without any labeled data. The most common applications are clustering (grouping data in categories based on their similarity) and dimensionality reduction (reducing the number of variables in a dataset while preserving its structure or distribution).
Deep Learning: Deep learning is a subset of machine learning that deals with algorithms called neural networks, particularly deep neural networks. These networks are capable of learning complex patterns from large amounts of data. “Deep” refers to the number of layers in the network; the more layers, the deeper the network. 
 
When you combine these two concepts, you get unsupervised deep learning, which involves using deep neural networks to learn from data that hasn’t been labeled. The above anomaly detection is one unique application that can be used in the industries.
 
 To watch the entire sharing on Youtube and we appreciate if you can follow us by subscribing to the channel. Thank you.

Joining our Free AIoT Workshop for Free!

We are offering a Free hours workshop ( valid till end of 2023) 

Scan the QR code below to register!

Industrial IoT in-band services for gateway management

Industrial IoT device in-band services for Terminal Management

In our 12th session of “Let’s Sembang AIoT,” we delve deep into the significance of Terminal Management Services for successful Industrial IoT (IIoT) deployment.
 
Given the diverse landscape of IIoT, edge gateways are strategically placed in remote locations, each with its unique connectivity—some leverage local broadband while others rely on mobile networks like 4G or LTE. Ensuring uninterrupted operations is paramount, and this is where Remote Management Services shine. They play a pivotal role in maintaining the optimal condition of these gateways and guaranteeing continuous, crucial data acquisition.
 
When deploying with Axiomtek’s IIoT edge gateway, Axiomtek provides two comprehensive in-band services. These services enable real-time monitoring of the gateway’s condition and health metrics, including CPU temperature, system temperature, power supply voltages to vital components like the CPU and chipset, and event logs for peripherals. This data is effortlessly captured using the Redfish agent, which offers REST-API in JSON format, or the Think Board agent that disseminates all pertinent details via the MQTT communication protocol to the Think Board IoT platform. Such capabilities accelerate the development and efficient management of your IIoT deployment.
 
During our session, CC Lee emphasized an intriguing point: by leveraging these pre-installed agents, the need for highly specialized programmers diminishes. With these tools, even those without deep programming expertise can facilitate IIoT deployment in a more cost-effective and expedient manner.

Live demo with Phyton coding to access the GPIO of the edge gateway

A live demonstration is done on the Redfish agent that perform the IO control on the in-built GPIO at the Axiomtek IoT edge gateway. 

Facial Recognition with Vision AI live Demo

 
During the second segment of our live stream, we delved into the realm of Vision AI, specifically focusing on facial recognition. This topic shed light on the potential applications of Vision AI in enhancing workplace safety. The principles behind facial recognition can be adapted using various AI models. Object detection methodologies can be integrated to establish an efficient workspace safety and security system, especially when combined with affordable standard IP cameras available in the market.
 
In this session, Kien Leong demonstrated how facial recognition AI algorithms are implemented. Our team members also participated, showcasing the capabilities of Vision AI in real-time.
 
Interested in learning more? Subscribe to our live channel and stay updated with our future sessions. Our motto is: “We Innovate. We Apply. We Share ( We Train).”
 
 
We value your feedback! Please comment if there are areas for improvement or any inaccuracies in our presentation. Thank you.

Subscribe to our Youtube channel if you find this relevant!

Jom lets sembang AIoT Youtube channel at AIoTmission