Blog

Any typical industry today comprises many processes with different complexity. These processes can be broadly divided into five logical levels. This categorization is done based on the nature and complexity of industrial processes. Although there is a lot of overlap between processes nowadays, the fundamental activities at each level remain the same. This can be visualized as a pyramid as shown in the image below. This is called as Industrial Automation Pyramid. Level 0: It gives a field-level view. This level directly works with machines operating on the floor shop. Sensors and Actuators are like the eyes and arms. One sensor just works with one variable. For example, a temperature sensor senses the temperature at one location. If there is a need to measure the temperature at multiple locations, many sensors have to be deployed. Similarly, one actuator controls one parameter alone. Level 1: It gives a machine-level view. A machine performs one unit of a task. For example, at a typical manufacturing shop floor, there are machines to perform tasks such as milling, grinding, drilling, cutting, etc. There are specialized controllers to control all functions of the machine as a single unit. These controllers internally use information captured by sensors and actuators to perform the task. In manual operation, an operator typically operates one machine and performs one task at a time. Level 2: It gives a process-level view. It’s an aggregate of tasks performed by individual machines at Level 1. Several machines performing their tasks in a sequence to achieve an output. For example, a block of steel will undergo several processes to take the shape of a gear. The outcome will be the final product or an unfinished component. Level 3: It gives a plant-level view. The typical plant consists of several processes such as raw material handling, machine job, paint shop, packaging, etc to convert raw material into a final product. In some cases, raw material could be components that are assembled into the final product. Level 4: It gives an enterprise-level view. It’s a top-level view that oversees and manages the complete business. Operations like manufacturing, sales, purchase, finance, and payroll come under this level. Flow of Information All levels are interconnected with the industrial communication system. When we go from level-0 to level-4, information gets aggregated. The level-0 and level-1 are time-critical activities. It’s a hard real-time system. Events and their response must happen in the allotted timespan. Response time is less than a second. Level 2 operates in a soft real-time manner. It has to complete under the allotted time slot but response time is not as stringent as level-0 or level-1. Processes involved in the lower three levels come under industrial automation. The top two levels come under information and communication technologies(ICT). There are many mature and robust applications, tools, and technologies to automate manual operations at all levels of any manufacturing industry. A discussion of individual technologies is out of the scope of this article. Roadmap for Automation Complete automation of any plant or factory or office cannot be done in one or two quarters. First, an automation roadmap has to be prepared. In the first stage, non-critical processes can be automated. Measure the benefit observed from automation. Then go for the next stage of automation. Automation includes training and skill enhancement of the existing workforce. Data collected from the bottom level is processed and analyzed by its next upper level. Processed information trickles up to the management. This is one of the benefits of automation. Accurate data from each level can be collected. This can be visualized and analyzed to derive business insight. Without basic automation, the Indian manufacturing industry can not adopt and benefit from advanced solutions such as Industry 4.0, predictive maintenance, and AI.

Data Acquisition devices are used to collect physical condition data. This data is used for IIoT, Industry 4.0, and compliance purposes. Physical condition data comes from multiple sources: Process data from plant and machinery: Temperature, Vibration, Pressure, Speed, System uptime amp; downtime, Fuel level, etc  Physical Environment: Noise, pollution level, Humidity, Temperature, Liquid level, Light intensity, etc The data is collected using various sensors. Industrial equipment such as SCADA, PLCs, energy meters, etc generates various data points. Data is received in both analog and digital formats. DAQs process collected data and send them to the computer system for industrial and scientific usage. Modern DAQs are full-fledged systems consisting of HW and software applications. The software includes a dashboard to visualize the data received from HW. Various communication interfaces are included to handle data from multiple sources. With automation and Industry 4.0 taking center stage in the current industrial setup, the role of DAQ has increased manifold. DAQ Design Consideration The design and architecture of DAQ differ depending upon the application. Some important considerations:Type of data being collected: Physical property of the system being measured.Sensors being attached to it: How many sensors of what type are being connected to DAQ? External or internal sensors, the power source of sensors.Storage Media: RAM, EEPROM, USB, amount of storage, memory management.Data communication: Wired or wireless communication, different communication protocolData processing or filtering: Does data need some local processing or filtering, the microcontroller used, etc.Power source: Battery, AC, or DC supply.Usage: Use cases. Based on this, the enclosure or external packaging of the DAQ will be designed.The DAQ manufacturers need to design the system taking the above points into consideration. DAQ Block Diagram The diagram below gives a high-level view of DAQ. Detailing each step of data acquisition electronics is out of the scope of this article. Electronic Components such as multiplexers, Sample-Hold, Anti-Aliasing Filters are used in various combinations depending upon the architecture of DAQ. Many of the components are not shown for the sake of simplicity. Analog signals can be used directly for any practical purposes. Signals captured by sensors from the physical system or machinery are often weak and full of noise. Signal processing techniques are used to process and make them useful. This is done with the following methods:Signal Processing or Conditioning: Performs amplification, isolation, filtering, and linearization on the input signal.Analog to Digital Conversion: Converts analog signal into digital format.DAQ must read the signal at the right sampling frequency. The incorrect sampling rate may introduce low-frequency errors which will distort the original signal. Controllers used in DAQs are either: Microcontrollers: Common Industrial and commercial use cases.FPGA: High-end use such as Defence and critical scientific use cases. Data Acquisition Setup Capturing industrial data is the first step in the automation process. Industries that are going for automation and Industry 4.0 must plan and decide their data strategy. Stakeholders must devote some time to identify and select critical data points. Once, data strategy is ready, set up the correct infrastructure to collect and acquire data points. Data has to be adequate and accurate. Data accuracy is not possible without the correct design and implementation of data acquisition infrastructure.  Summary A wide variety of DAQs is available catering to all types of requirements. Industries can procure standard DAQs off-the-shelves which are adequate for most of the common use cases. DAQ manufacturers can manufacture customized DAQs as per the applications requirement. A well-planned data strategy along with robust and accurate data acquisition infrastructure will help industries achieve their Industry 4.0 goals.