Robotic arms busy in a car plant

Smartening up the factory floor

Smart automated factories are now crucial in how processes operate, and it is sensors that drive the change.

The boom in connectivity that transformed the world of consumer electronics is rapidly becoming a driving force in industrial automation too. Nowadays, individuals expect instant access to data, up-to-date status reports on the news and weather, and links between their gadgets, so much so that the same expectations now apply in the factory, to make automation smart and part of a connected network.

Intelligent industrial automation has become a priority among companies, and in particular smart sensors are the backbone of the whole intelligent network. As factories adopt modern automation and production systems, embedded sensors are used for monitoring conditions such as temperature, or to access the operation of equipment, which makes it possible to reduce downtime and repair costs.

The future factory is here, and as Germany leads the fourth smart factory revolution (see box) it continues to evolve.

Smart factories have three life-lines, says Gavin Stoppel, product applications manager in smart networks infrastructure at Harting: power, data and signal. Power provides the energy to the manufacturing line, data provides feedback to the central control system, and the signal integrates the parts of the machine around the network. Smart sensors need to be integrated with controller processing data which it transmits to a control system using real-time data output interfaces, such as Ethernet and EtherCAT. Smart sensors handle signals using Wi-Fi, Bluetooth or industrial communications protocols such as Ethernet, ProfiBus or TCP/IP.

Stoppel believes the seamless exchange of data is an advantage on many levels. 'Knowing where everything is – product, machine, tools and people, makes for just-in-time manufacturing, with everything in the right place at the right time.'

Sensors can create a flexible manufacturing environment, changing production, for example, with a simple change of the sensor parameters. Equally, sensors can check tools and equipment and send an alert before a part is due to fail, thus reducing downtime and losses.

'There is a proliferation of sensors and actuators that reside at the 'edge' of IoT (Internet of Things) systems, where large amounts of data from our physical world (something we call big analogue data) is converted to a digital signal,' says Rahman Jamal, global technology and marketing director at National Instruments. 'By performing analysis and control on sensors and actuators, early in the data life cycle, businesses can derive value faster.'

He adds that hardware and software platforms that acquire information from multiple sources, and measurement types that can perform analytics closer to the sensor, can minimise data overload to central servers. 'The idea of a smarter world where systems with sensors and local processing are connected to share information is taking hold in every single industry. These processing systems will be connected on a global scale with users and each other to help users make more informed decisions. This will extend beyond the industrial setting, to include smart cities and the smart grids.'

What's the difference?

Within the smart factory two groups of sensors are commonly used: electromagnetic and optical devices.

Optical sensors use a laser diode and photo elements to measure the distance between it and the object. Short laser line sensors are used on metallic surfaces, while the longer laser line of laser scanners is used to detect multi-dimensional profiles of objects. Both are commonly used on automotive production lines.

'Smart optical sensors have an integrated controller for the sensor to perform tasks,' says Chris Jones, managing director at Micro-Epsilon UK. He adds that other optical sensors include confocal-chromatic measurement systems, with Ethernet or EtherCAT network interfaces. These are used to measure the thickness of transparent, multi-layer objects, as well as for distance and displacement measurement.

'The user benefits from a tiny measurement spot and nanometre resolution,' Jones explains. Confocal sensors are used in quality control of high-tech production lines as the fastest can achieve speeds of 10kHz using an LED light source and 70kHz using a xenon light source.

Other sensors are based on electromagnetism. Magnetic technology sensors have been used for at least 20 years, before the idea of the smart factory was conceived, says Hiroshi Dan, product manager in sensor products at Murata Electronics Europe. Today, they control machine parts, such as the angle of tilt in robotic arms. Rotary position sensors measure the angle of rotation using a microcontroller inside the equipment. Solid-state, non-contact magnetic technology is essentially wear-free, and therefore reduces maintenance.

Magnetic strengths can weaken over time, but sensor manufacturer Balluff has introduced magnetic field sensors that work with field strengths down to 15 gauss, meaning that operation is not impaired even when the magnet is weakened by age or heat.

'Smart sensors are used to transmit and receive data, and to act in a particular way,' explains Harting's Stoppel. He divides them into active and passive versions. An active sensor has its own power source, a battery for example, whereas a passive sensor responds to a request for a status and transmits that data to a central point. An RFID transponder is an example of a passive sensor.

'If there are people around the plant, an active sensor can track movement,' explains Stoppel, 'whereas a passive sensor is better for items that need to be located, for example, to update customers on their delivery or for factory workflows.'

Sensors cut human error

Local processing and communications to make machines smarter is the intelligence National Instruments added to the tools and shopfloor systems at aircraft manufacturer Airbus. Jamal explains how the production process was simplified and production efficiency was improved, by managing and checking tasks.

One example of this is a sub-assembly of an aeroplane with approximately 400,000 points that are required to be tightened, requiring over 1,100 basic tightening tools. The operator has to ensure the proper torque law settings for each location using the correct tool.

'Human error adds a lot of risk to the production; even a single location being tightened down incorrectly could cost hundreds of thousands of dollars,' says Jamal. 'A smart tightening tool understands which task the operator is about to perform, using vision to process its surroundings and automatically set the torque. The device can record the outcome of the task in a central database to ensure the location was set properly.

'With the central MES (manufacturing execution system) database and the distributed intelligence of the devices, production managers can precisely pinpoint the procedures and processes that need to be reviewed during quality control and certification.'

Further information

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles