How to build a smart factoryAnand HarigopalSuyambuAmnbihaiSmart factory After the industrial revolution, automation in production has achieved a tremendous growth. The next giant step in production will be construction of smart factory which uses the concept of industry 4.0 and artificial intelligence. Smart factories work on the principle of interaction between the physical object and the control software which are interlinked by Internet of Things (IoT), where the information technology is interlinked with the operational technology, right from the start of inventory control till the delivery of the fully finished object. In some necessary places the robot interacts with the human. All the data are digitalized and stored in the database for future reference. These factories are so flexible that the production processes could be varied automatically according to the product being produced which is controlled by the centralized cloud network. By this highly standardized production process, highly precise and high-quality products can also be delivered to the customer.Digital supply network To construct a smart factory a central hub of digital supply network is necessary. Unlike the traditional production supply chain, the centralized digital supply network can interact between various supply chain systems and the production process can be customized. This helps in achieving a highly flexible production and supply chain system. All the information’s are collected, processed and transmitted by this digital core, which decides the sequence of operations to be carried out and hence the line production system can be eliminated. Smart factory features:People as key players: In manual and semiautomatic production process the split of human resource and utilization of the machineries determines the efficiency of the process. In a smart factory each employee’s skill set is fed in the central digital core system and once the employee enters the work place the system identifies the worker (might be with a barcode or magnetic strips) and allots work according to his skills. It also monitors the work done by the worker and provides him with appropriate tools. Ergonomics of the worker is also monitored and provides recommendations to ease of work. Distributed intelligence: The central control system is then decentralized, and the information are send to the sub artificial intelligence, whichsplits the work and get it done. The main central system sends the product specification to the subsystem. The subsystem then makes autonomous decision and starts the process by controlling the drives, sequence and operations. The artificial intelligence continuously monitors the changing requirements and working conditions and adopts itself according to the atmosphere.Open standards: A smart factory must always have fixed and open standards. These standards should be the same all over the process and system. Once the standards are decided, there should be no deviation from the standards to avoid malfunctioning or production of low quality products. The standards of the Ethernet and connections should also be standardized. Two machines with two different standards will interact between them which are not understandable by each other. Virtual reality: The complete elements and process of the smart factory are in virtual reality time representation. All the virtual parts interact with the physical components to perform the operation. For instance, the central hub collects all the information from the surrounding through software and sensors. These data are then send to the hub system through Ethernet and processed. The feedback is send to the control system and the above procedure repeats without any intervention of humans. Lifecycle management: As the computer system decides the process to be carried out, it analyses with various data in nature and finds the easiest and the efficient way to solve it. Hence the development or planning time, setting time, process time and cost can also be reduced significantly. The customer can also easily track his product when it is getting manufactured. Variance of the products been produced can have good scope with minimal non-value-added time. Smart factory characteristics:Connected: The main characteristic of a smart factory is its connectivity of the material with the process, relationship between the departments of the factory, interlink between the customer, supplier and production process along with thecapability of Artificial Intelligence to take decisions in real time. Sensors should be setup in all possible sub system to extract data and keep the system updated. The system also retrieves the historic data of the past and relates them with the current data. It should also retrieve information from the customer and the supplier end to create a highly efficient supply chain and network efficiency.Optimized: The factory should be designed in such a way that all the process should be optimized to have its least process time. Automation should be given higher importance and human interaction should be reduced as much as possible. In the upcoming years energy consumption is also more important, so renewable energy sources should be preferred. The other two things to be noticed is quality and wastage. Highly precise and high-quality products should be produced with minimum wastage as possible. All these optimizations will automatically reduce the cost of the product. Transparent: All the real-time data should be transparent for decision making. When the process and data are transparent, the faults can be found easily and rectified. It is possible for the system to provide with notifications and alerts. It also helps the customers to track their products and monitor the process.Proactive: Precautions should be made before the issue or failure occurs. Due to the transparency of the process, decisions can be made before the fault. It also includes monitoring the restocking, inventory, quality issues, maintenance and safety. The historic data’s can also be used as benchmark for maintenance and prevent breakdown. This helps in reducing the down time of machines.Agile: Due to flexibility of the smart factory, scheduling of the operations and material requirements are planned automatically and executed. Then the data are cross checked with data of real time, and the system updates automatically. Advanced smart factories should be capable to plan the requirements according to the change of product. The best possible way should be selected by the system, so that the setup time, scheduling time and changeover time can be optimized.Process within a smart factory:ManufacturingOperations Additive manufacturing helps in producing rapid prototype and low volume products.Advanced scheduling and planningreduces waste and cycle time using real time inventory and production data.Autonomous robots can be used in recurring process which is repeated again and again with high accuracy.Digital twin can be used to integrate the predictive analysis by digitalization of operation.Warehouse operations Augmented reality assists humans in picking and positioning the articles.Autonomous robots do all warehouse operations with the help of artificial intelligence. Inventory tracking Sensors can be used to know the real-time position of the raw material, their storage in warehouse, their stock, current work status and tracking of finished products.Analytics to reduce the inventory, to monitor them and automatically order them when there is demand.Quality Optical based analysis systems to be used to find defects.Equipment monitoring to find the defects in the machine which produces the product.Special sensors which identifies the flaws and does the quality check.Maintenance Augmented reality, the robotic system which helps the maintenance people in repairing and maintaining the equipment.Sensors to detect the malfunction of the machine and conduct the predictive maintenance. Environmental, safety and health Sensors to monitor the health of the person and send notifications in times of emergency or saturation limits.Sensors to predict the environmental changes and take precautions before the system gets affected. Sensors to monitor the machines to prevent accidents due to machine failures.Key physical components of a smart factory:Sensors: Sensors are detection devices which monitors the device around the clock and provides digital data for the system. Countless sensors are available in market, but the sensors characterized according to the application should be used to monitor the process.• Thermal sensor (thermometer, thermocouple, pyrometer, net radiometer, etc.)• Pressure sensor (barometer, pressure gauge, piezometer, ionization gauge, etc.)• Force sensor (piezoelectric sensor, strain gauge, force gauge, torque sensor, etc.)• optical sensors (photo detector, photo diode, optical position sensor, etc.)• position sensors (gravimeter, odometer, tilt sensor, tachometer, flex sensor, etc.)• navigation sensors (air speed indicator, altimeter, variometer, turn coordinator, etc.)• Ionizing radiation sensor (cloud chamber, Geiger counter, neutron detection, etc.)• Flow sensor (air flow meter, gas meter, mass flow sensor, anemometer, etc.)• Environmental sensors (frequency domain sensor, rain sensor, soil moisture sensor, etc.)• Electric sensor (current sensor, voltage sensor, magnetometer, electron multiplier, etc.)• Chemical sensor (pH sensor, hydrogen sensor, optode, ozone monitor, etc.)• Sound sensor (geophone, seismometer, hydrophone, microphone, etc.)Valves and actuators: Valves and actuators are the basic electro mechanical components which controls the speed of the process. These basic components of automation are controlled by software, in which the process requirements are fed.Example: ball valve, butterfly valve, clapper valve, check valve, choke valve, pinch valve, piston valve, plug valve, poppet valve, safety valve, etc.PLC system: The programable logic circuit is the base of automation. The information collected by the sensors are read by the PLC and analyzed. The sequence of events can also be programmed in the PLC, according to which the process is carried out. The actuators and valves are also controlled by the signals from the PLC circuit. The most hazardous and complicated works should be replaced by automated devices controlled by PLC’s. PLC’s are used ino Assembly lineso Highly reliable activitieso High precision activitieso Robotic deviceso Process controlsSCADA: Supervisory control and data acquisition can be used to control the activities or progress of the activities. This could also be used to control the efficiency of the systems. Once installed to the main system, it controls the PLC’s and retrieves data from them and commands the PLC’s for the further actions. SCADA is the backbone for the automation and control systems in all the smart factories. The main applications of SCADA s in the fields of oil, energy, road and transport, power and grid, water and resources, manufacturing, recycling, air and sea, etc.OPC server: The object linking and embedding for process control is the bridge between the physical component and the control system. The OPC server translates the data collected from the hardware to through PLC into the OPC control. The OPC historic data access is used to access the data from the past and analyze it to know if it could be used as a bench mark for the new process. IoT gateway: The Internet of Things is the gateway between the cloud of data and the accessories in the real world. The data transfer is bidirectional, and it can send thousands of information per second. The gateway also has the capacity to process the data before it is send to the cloud or to the component. Due to bidirectional data transfer, the data security is high.AR/VR: The Augmented reality and the virtual reality could be used to troubleshoot faults and failures in minimum time. The other features are• Increase efficiency by increasing the workers safety.• The virtual reality collects information during the virtual reality experience and computes it with the historic data saved in the database. • Development cycle reduces with the augmented reality.• Planning time can be reduced using augmented reality and high-quality planning can be achieved. • Production system need not be modeled full.Steps tobuild a smart factory:Step 1: Create a concept of the smart factory in a manageable environment. Planning should be done from the base like, the product to be produced, followed by the machine and machine tools required to produce the product, and then the dimension of the factory required to hold all these machineries along with the storage and buffer facilities. Step 2: Scale the basic concept into a prototype by increasing the performance of the tools and machineries. This prototype should hold all the basic information required to construct the factory. This prototype can be made physical with trial methods or in virtual by simulation software. Step 3: Once the prototype succeeds, the additional assets should be added. This include addition of all supporting process. The connections between the sequential operations should be validated.Step 4: Implement the production line by improving the performance of the dependent processes. The implementation should be done such that the setup time, waiting time, and the process time is reduced. Step 5:Improve the factory by the better utilization of the resources and assets. The major thing that should be considered is lowering the inventory and stock in the mean time utilizing the machine to its optimum level.Step 6: Networking should be done by linking all the process to the central hub. The supply chain should also be connected to the central system. The connections may be through wires or through Internet. In this case every machine will have a transmitter and a receiver that would help to code and decode the commands and messages. Advantages:?