From the first days of the Industrial Revolution, the manufacturing sector has been subject to a swathe of technology trends that have transformed the ways goods were produced. In the first wave of change, cottage industry workers were replaced by machines in factories, transforming the nature of manufacturing through automation.
In the 20th century, mass production was ushered in, while in more recent times, robotics has powered further transformations to the assembly line. Now in the 21st century, manufacturing is undergoing a revolution again, this time through digitisation and the rise of new technologies that improve productivity, processes and drive new ways of thinking.
Here, we take a look at some the emerging technologies now shaking up this industry sector.
Productive 3D printing
One of the most impactful technologies in manufacturing is 3D printing, which can cut down production times and give more flexibility in the design of products. Keech – which manufactures parts mostly for mining, rail, defence and agriculture companies – has cut the time to manufacture several products from two years down to six months thanks to 3D printing.
Herbert Hermens, CEO of Keech, says 3D printing allows the company to design parts more accurately as well as employ more complex geometries, thereby cutting out the rigorous, time consuming casting processes its engineers usually go through.
“We are able to translate the theoretical into a practical application by making the models overnight. It allows us to almost overnight go to customers and ask them what they think of the design and then come back and make those modifications,” he says.
“When you can cut down a design cycle, as we have in some cases down to a quarter of what it was, all of a sudden you are saving hundreds of thousands of dollars in the design process.”
Having more design flexibility with 3D printing has also allowed Keech to broaden and enhance its product range. Recently, the company engaged with CSIRO as a commercial arm to start printing medical products.
“It is absolutely being able to design this to the end-user requirement,” Hermens says. “We can change designs so that our productivity is much higher and as a consequence, the customer finds much better value in our products.”
In addition, thanks to these clever innovations around design, Keech’s engineers found a way to reduce the amount of material used to manufacture a hinge while keeping the same level of integrity. The engineers cut a third of the weight of a 3D product, dropping production costs by 10 to 15 per cent.
However, the success of 3D printing at Keech is just as much about its engineers as it is the technology itself. Understanding how materials behave is crucial to getting the most out of the 3D printers, Hermens says.
“You can print everything from ceramics, to different metals, to different plastics, and they all behave differently. Because we understand material and material behaviour… we can test the veracity of those materials using a gauge against the final material,” he says.
“For example, if we are making a steel product, we don’t have to immediately go to a steel prototype. We can first do this in ABS plastic and conduct a theoretical test and relate that to how steel would react in the same circumstance. So while we still do final testing in the final product, and around the final size of the product, we cut down the amount of testing dramatically.”
Keech invested significant time in training engineers on how to best use the technology, and to learn about the potential of different materials within a broad canvas of possibilities.
Speed to market is key to survival in the industry, Hermens says, which is what 3D printing technology in manufacturing is all about at the end of the day.
“We need to start running ahead of the market, so we need to develop new products much faster and expand the product range,” he says. “I call it mass production to the individual. You can still maintain a mass production approach but you can make modifications much cheaper now. In automotive in Germany, for example, they are using 3D printing to individualise cars, already.”
Smart sensors and data
Monitoring equipment to prevent failure and adverse effects on business, as well asset optimisation, is a top focus for many manufacturers these days thanks to the rise of the Internet of Things. GE, which manufactures large jet engines and locomotives, is all about doing that, using embedded sensors and analysing the data they transmit over the Internet.
A GE jet engine collects 5000 data points a second or a terabyte a day, while a locomotive generates 9 million data points every hour.
“We are collecting data from machines in real time, transmitting vast amounts of data over the Internet, analysing it in real time, sending it back to the machines and making a change dynamically,” says Mark Sheppard, CIO at GE Australia.
“This could alert the owner that some characteristics of that equipment might be performing in an unusual way. For example, the oil pressure or temperature was too high, or something was operating out of sync. It wouldn’t necessarily mean the equipment was going to fail straight away, but it might be an indicator that it might fail soon.”
The GE jet engines are able to communicate with ground operators while in flight if they are likely to need repair or maintenance work in the near future. So by the time the aircraft lands, staff can get onto it straight away and not have it manifest into a sudden, serious issue. It means they can also plan ahead for downtime.
With locomotives, GE also keeps track of the condition of the equipment through sensors, as well as optimises the locomotives as part of a whole system to make savings on fuel burn. According to GE’s calculations, a 1 per cent improvement from monitoring and optimisation could result in $30 billion in savings in aviation and $27 billion in rail/locomotives over 15 years.
However, Wi-Fi connectivity in remote areas where some of this equipment is used, such as in mining, along with interoperability, are constant challenges, Sheppard says.
“The locomotive might be travelling across the Pilbara, where we might not have access to 4G data networks. So we are working with clever techniques such as ‘store and forward’, where we will collect the data as the locomotive travels along,” Sheppard explains. “When it comes into an area of connectivity and gets a Wi-Fi signal or a cellular signal, the data is uploaded.”
The sensors also prioritise what to transmit depending on how strong the Internet connection is. For example, if the locomotive is in a satellite area, it would first send any data out that could indicate an issue about to occur. When it reaches a Wi-Fi signal, it then sends the remaining data.
GE uses its in-house developed Predix platform to communicate with other manufacturing equipment, enabling interoperability. “It acts as a common way, a common interchange of collecting that data and do the analytics inside Predix,” Sheppard explains.
Factories are risky environments when it comes to safety, and things can get rather complicated when staff need to work alongside machines. That’s where human-friendly industrial robots come in.
Prysm Industries, an injection moulding company, is using UR5 robots from Universal Robotics for container labelling tasks where humans work alongside the machines. “Because they are round containers that have to be on a certain part of the product, we can’t just run it through a generic inline labeller,” explains Matthew Murphy, production manager of Prysm Industries.
“That’s where the UR5 comes in handy. It can pick up the part on the conveyor, put it into position in front of the labeller, tell the labeller to label, rotate it around so it’s in a perfect position every time, and then put it on an outgoing table for the operator to put a lid on and start packing into boxes.”
A staff member can safely stand in close proximity to the robot, allowing him or her to engage in other activities while the machine is operating.
“It just enables the operator to not really be stuck at the machine, and it enables them to walk around and get more lids if they need, place the boxes on the pallet, while the machine is running,” Murphy says.
When a person stand unusually close to the robot, it’ll stop completely, he continues. In the worst case scenario where it hits a human, it’s designed specifically to avoid forceful impact. “I haven’t had it hit an operator yet, but if it does it really is quite gentle,” Murphy says.
Prysm’s objective is not about replacing staff, but freeing them up to operate other machines and do higher-level tasks, leaving mind-numbing and repetitive tasks to the robots.
“Also, labelling has to be in a set position every time, has to be applied without creases or air bubbles - the robot does it the same time, every time, and a lot better than a human can,” Murphy adds.
However, there is a minor restriction with the UR5 in that it only has a reach of about a metre so it can only handle objects within about a metre radius.
If there is one technology disrupting not just car manufacturing but the nature of driving itself, it’s the driverless car. Having driverless cars optimise themselves to pick up and drop people off upon request through an app, so they never have to buy and own their own car, is not far-fetched.
GoGet in Australia, a car sharing service company, is working towards this ambition by engineering its own driverless car. It currently has a test car and is further developing the technology to roll out in future.
GoGet Carshare CEO, Tristan Sender, says the car would work out the best way to pick up and drop off customers. “It would work out through algorithms to make sure they are being as efficient as possible with where they go and who they pick up,” he says.
“The car would already probably have who it would pick up next before you even get out of the vehicle.”
GoGet is investing in the driverless car as it sees this as the next step in car sharing.
“It’s a more efficient version of sharing. So we may provide the self-driving car service to people just like we currently provide car share service to people,” Sender says. “You’ll probably see large savings in terms of cost as well, because the level of efficiency of these vehicles.”
The South Australian Government announced in February that it will reform legislation to allow for driverless cars in the future, with trials to take place on public roads in November. Sender says this opens the door to innovation around car services. “It allows for new things to be developed that will benefit society in the long term,” Sender claims.
“What this is going to do is make other transport authorities think more about it,” adds deputy director of University of NSW's Research Centre for Integrated Transport Innovation (rCITI), Vinayak Dixit, who is developing GoGet’s driverless car.
The car takes in real-time feeds from what’s happening around it through sensors and its perception system. The next step is having vehicle-to-vehicle and vehicle-to-infrastructure communication.
“Real-time information about vehicles and their intentions of travelling through a certain route would actually help optimise on travel time and emissions,” Dixit says. “If you know 3000 vehicles are going towards the city and taking a particular road, the car could actually make a decision that the road is going to be congested in the next five minutes, and will reroute.”
However, until then, driverless cars would still have to operate with traditional cars and human drivers on the road.
“One of the biggest challenges is intent. As you come to an intersection and you are turning right, you really need to know the intent of the person/car in front of you – if they are turning or going straight,” Dixit says. “That’s the hardest part to address. If you don’t have communication between the vehicles or between the infrastructure. It makes it hard for these vehicles to be driving on city streets.
“Recognising their intent by even small adjustments to the drive, subtle movements, means you start compensating for other drivers.”