Balancing the benefits of automation

From building luxury cars to making cricket balls, from running postal services to producing factory equipment, automation is the order of the day – but the solution has to be tailored to the business.

Automation and the human touch: Lexus

Volume car makers are highly automated, but even luxury marques whose value depends on manual build and hand-crafted interiors rely on automated processes for their vehicles’ development, manufacture, and drive controls. The distinguishing features and customised finish that make a car individual, however, are still rendered by the hands of super-experienced craftspeople with skills unmatched by machines.

Lexus runs some of the most advanced vehicle manufacturing in the automotive sector, but in many areas of its design and finishing the human touch prevails – such as in the development of car shapes. Digital rendering enables designers to experiment with shape, line, and surface, says Lexus chief designer Tadao Mori, but no technology has yet been able to replace traditional clay modelling when it comes to the evaluation and refinement of a full-sized design prototype.

“We tried many digital modelling technologies, but in the end, we always came back to clay modelling by hand,” Mori explains. “Computer-aided design [CAD] systems have their benefits, but it can be flat and a bit cold. Clay allows us to make the designs look and feel more ‘human’.”

With over 10 years experience, Yoshitaka Yanamoto belongs to a clay-​modelling team that aims to craft each curve and contour of Lexus cars to evoke an ‘emotional response’ from prospective buyers. Yanamoto uses tools to shape one side of the model in clay according to a 3D design drawing.

That side is then translated into a CAD system, and mirrored to create the opposing side. The modelling process can take up to four months, as design and engineering departments collaborate to take a vehicle into initial build and production.

Use of non-automated skills for many aspects of car production is due to Lexus’s cultural commitment to ‘takumi’ – artisans dedicated to honing their engineering skills toward a concept of perfection. At Lexus’s Tahara plant in Japan, there are 10 takumi master craftsmen in the areas of engine casting, engine machining, engine assembly, moulded parts/plastics, plastic painting, stamping, welding, body manufacturing, painting and assembly, and inspection. Each leader is responsible for teaching technical skills and the development of production and audit processes in their section.

In charge of final vehicle inspection, Katsuaki Suganuma and his team are trained to inspect painted surfaces and panel gaps using their hands and eyes only. Suganuma has spent 60,000+ hours using his highly tuned senses to detect the minutest imperfections throughout each Lexus car at the final stages of its production. “It is about how a car feels when you touch it, or the experience of sitting inside it,” he says. “Robots will not be able to replicate that.”

Bowled over by automation: Kookaburra Sport

In manufacturing, the job of loading and unloading machines with parts or materials has traditionally been done by humans. The production of cricket balls at sports equipment maker Kookaburra Sport was a good example: workers at the company’s Australian factory would place covers – the outside leather – onto trays after they had been pressed into shape to form half a cricket ball.
These trays were then placed into a rack and moved to the trimming process, where workers would pick and place each cover into a trimming machine, which used a blade to remove excess leather from the ball.

Working with tech partner MAP Services, Kookaburra wanted to replace this manual process with an automated robotic system for loading and unloading the trimmer, including a tray-dispensing system. The plan was to place the full trays of cricket ball covers into the dispenser, feeding trays automatically to a TM Robotics Shibaura Machine TVL700 robot integrated with MAP Services’ vision system, to load and unload the trimming machine. This would require only one operator to load a set of trays every 20-25 minutes.

MAP designed an automated racking system, where six trays carrying about 85 cricket ball covers are placed onto a conveyor. This part of the application usually scatters the covers in random locations – so the vision system has to guide the robot to where each cover is and how to pick it up. The computer-aided detection-based system’s cameras continuously capture images: using software intelligence, the system then processes these images and identifies the precise position of each cricket ball cover.

Next, the system determines the most logical order to pick up each cover. The robot takes the cover and puts it in the trimming machine, after which empty trays are fed into a bin for collection.

Overall productivity gains provided by Kookaburra’s new machine-tending system are equivalent to 90 extra minutes each day – or a 15 per cent increase on the previous procedure. This gain is not necessarily due to increased speed of production, but to constancy, according to Kookaburra. Where human workers tire and slow down or take breaks, the robots continue to work at pace.

Ultimately, the deciding factor in Kookaburra’s choice to automate some of its production processes was the ability to extend the working day, but the new system has also resulted in a reduction in the number of balls that do not pass quality inspection standards.
With an operator required only to load trays every 20-25 minutes, other workers have been upskilled and re-deployed in Kookaburra Sport. Rather than hiring new employees to operate the robotic system, Kookaburra chose to upskill existing ones, training them on how to stop, start and recover the automated process.

The robots stop working when the last person leaves for the day. As a result, production lasts longer each day, while labour costs remain the same. This is before operational expenditure is factored into the total cost of ownership (TCO).

“Many first-time buyers of industrial automation consider only the price tag of the robot they wish to buy, but this is far from the TCO of an automated solution,” says MAP services director Shane Gallagher. “Other important factors include installation, labour, energy, materials, ongoing maintenance and any peripheral technology required. These costs can vary wildly between businesses, primarily depending on size and industry sector. By comparing productivity gains and reduced labour expenses against the cost factors previously mentioned, businesses can work out whether automation is the right choice, right now.”

Pushing the automation envelope: Royal Mail Group

Multinational postal service and courier company Royal Mail Group has two main businesses, Royal Mail and General Logistics Systems, and employs more than 158,000 people globally. It delivers around 14 billion letters and 1.8 billion parcels per year. Like any commercial entity, Royal Mail looks for ways to bring efficiencies to its operations – an imperative that led to its transition to Robotic Process Automation (RPA) for its central business operations.

RPA is software that makes it possible to build, deploy and manage software robots that emulate human actions when interacting with digital systems and applications. Like people, software robots can do a range of tasks – such as understand what’s on a display screen, complete keystroke sequences, navigate systems, identify and extract data, and other defined actions.

Working with RPA specialist UiPath, in 2017 Royal Mail began an ongoing business operations transformation programme that started with process simplification and automation. Since then, UiPath’s RPA solutions have been introduced to six areas of Royal Mail’s business: operations, customer experience, finance, HR, commercial and fleet.

The initial ‘Proof of Value’ pilot project showed that each robot saves the time equivalent of two employees. However, according to Michael Hitchins, RPA programme lead at Royal Mail, the pilot robots also helped improve the quality of the processes they automated.
In Royal Mail’s operations team, for instance, software robots were used to change the way delivery office managers across 1,250 locations completed tasks. “Previously, when they started work at 6am, they had to access six systems, extracting data needed to brief their staff at 7am. Each manager would do this in their own way,” Hitchins recalls. “It took time, impacted productivity, and delayed them speaking with their team. We introduced unattended software robots to extract data automatically, presenting it to managers in a standard format, allowing them and their teams to focus on their operational tasks, rather than routine admin.”

In the finance department, robots were used to improve account handling. “Royal Mail has a huge number of customers. If one of them doesn’t pay on time, their account has to be suspended,” says Hitchins. “Once their payment is received, however, the account must be reinstated. Doing this manually took herculean effort and led to some accounts not being suspended or reinstated on time – with a resultant loss of revenue. Robots now manage this process in real time, with no revenue loss.”

More than 80 automated processes have been implemented across the business. As the automation programme rolled out, Royal Mail employees began to notice benefits in terms of cost savings, productivity and employee satisfaction, Hitchins says.

Hitchins and his team had originally aimed to deliver £13.7m in financial benefits. So far, they’ve delivered £55m – an ROI over 300 per cent above expectations, he reports. This comprises £4.8m in employee time saved, £22.4m in additional revenue, £17.5m in cost reduction/avoidance, and £10.3m in capacity creation. At the same time, approximately 662,000 hours of time have been ‘handed back’ to employees.

“After introducing automation, the results we got, such as cost savings, cost avoidance, revenue generation and freeing-up valuable resource capacity, were not really unexpected,” Hitchins says. “People tend to focus on cost reduction only, but by doing this they’re missing out on many opportunities. For example, there’s been improved engagement from people using software robots, as they ‘took the robot out of the person’, so to speak.”

Fully automated factory: Schneider Electric and AVEVA

Energy management and automation solutions vendor Schneider Electric and industrial software specialist AVEVA have digitally upgraded their factory in Flint, Wales, using the companies’ own automation solutions. The project, which started in late 2018 at an eventual cost of more than £300,000, will see the site become a fully automated smart factory making Schneider’s cable management and trunking product lines.

“From a production and maintenance perspective, the Flint factory used to be digitally blind,” says Mark Yeeles, VP of industrial automation at Schneider Electric. “The transformation project primarily involved a full retrofit of the existing factory to maximise efficiency and minimise downtime.”

The Flint project leveraged several of Schneider and AVEVA’s own digital transformation and operations solutions. The hardware solutions deployed were supported by the Schneider multifunctional EcoStruxure interoperable systems management architecture and platform. It provides the Flint factory’s managers with a suite of digital tools to control and optimise the facility’s operations.

The project has brought quantifiable ROI and ongoing cost efficiencies, says Yeeles: “Our investment was paid back within 12 months, due to productivity and a range of efficiency-based savings. Maintenance costs saw a 15 per cent reduction – about £91,000 – and we saved £14,000 due to the installation of intelligent lighting.”

Production efficiency has also been raised. Despite a 20 per cent decrease in demand in 2020 due to the Covid-19 pandemic, Flint’s operational efficiency improved by 5 per cent compared to 2019 levels and its productivity increased by 4 per cent – which equates to £240,000 in savings, Yeeles says.

“Transforming the factory also brought sustainability benefits,” says Yeeles. “Water consumption was cut by 50 per cent, from 4,228m³ in 2019 to 2,138m³ in 2020, with projections down to 960m³ by the close of 2021. Improvements to energy efficiency associated with the installation of variable speed drives, smart panels, and other technological solutions led to a 15 per cent decrease in electricity (kWh) usage compared to 2019 levels.”

Beyond the hardware and software introduced to the site, staff were made more aware of areas of inefficiency and enabled to take ownership of issues. They have access to dashboards, as well as the ability to create alerts to notify them of any changes within the plant. The newly empowered staff includes people who were previously employed on non-automated systems at Flint, and who have now been retrained and reassigned to digitalised jobs.

“Staff were able to learn new skills, develop in their careers,” Yeeles reports. “The focus has been on upskilling staff and allowing them to take ownership of the digital processes.”