Saturday 18th Nov 2017 - Logistics Manager

Automatically better?

Automation is the way forward for many logistics operations, but is there such a thing as a low-cost automation strategy. Sam Tulip investigates…

This article first appeared in Logistics & Supply Chain, June 2017.

Rapid change across different industries from manufacturing and retail to healthcare, is a reality. Ashokkumar Jawaharlal Bafana, managing consultant, travel & transport at Wipro, cites double digit growth in e-commerce with all the increased consumer expectations on speed of delivery, order accuracy, transparency, and reduced (or zero) shipping cost; regulatory requirements, for example on visibility in healthcare supply chains; and labour shortages and costs.

“To overcome these changes, the logistics and transport sector is being leveraged by different industries. Logistics players are under continuous pressure to reduce supply chain costs, increase operational efficiency, but maintain service quality.

“Is automation within the warehouse one of the answers to handling volume surges, speed, transparency, and labour shortages to stay relevant and competitive? Does automation help increase throughput or increase order accuracy? Does it reduce the impact of labour shortages?”, he asks.

Most countries, says Jack Kuijpers, vice president North West Europe at Witron, are facing demographic problems with the availability of labour, compounded by stiffer rules and regulations. “Older people can’t do heavy work after 40 years with the same efficiency, while younger people aren’t interested in this kind of work. Minimum wages are driving up labour costs and in Denmark, for example, there are rules on the maximum tonnage that can be handled in a shift, so an operator may earn eight hours’ money for three hours’ work. In the UK a lot of physical labour is carried out by people from Eastern Europe and it is not clear how this will continue after Brexit”.

Many companies are realising that they aren’t making money out of e-commerce home deliveries, especially with the cost of returns. “Companies are having to rethink their business models, but there is a lot of uncertainty, which we have to take into account, and not every business is developing in the same direction. Nothing is as flexible as a shed with a lot of hands, but there is a price. We have to try to predict the future, but organise our automation concepts around flexibility. Automation should be able to cope with potential changes over a generation”.

Warehouse automation is indeed increasingly attractive: in some cases, says Steve Richmond, systems & projects director at Jungheinrich UK, it is essential. “There are parts of the UK where the labour resource just isn’t available to provide the hundreds of operators and drivers needed for a large manual DC. Equally, land is scarce, leading to a requirement for high bay operations. The cost of automation has reduced in relative terms in recent years while land and labour costs keep rising, so there is a return on investment argument for automation in operations that ten years ago would not have been considered”.

But, says Richmond: “There is a common misconception, dating from the ‘80s and ‘90s, that automation is expensive and inflexible. But there is now a more appropriate use of technology, combining manual, semi- and fully-automated elements. Firms do not have to automate their full operation, as long as the manual elements are designed into the internal supply chain process”.

Francesco Papale, VNA & WH system specialist for Hyster-Yale Group, says: “Big players like Amazon are changing the rules. Flexibility is the basis of their structure – they can change completely in just a year. We have to cope by creating trucks and automation that can change at the same speed, and create solutions that smaller companies can use, because they must follow the same logic just to survive”.

Forecasts
When Jungheinrich designs a solution for a customer, says Richmond, there are three essentials. “We have to understand the supply chain process; we have to understand the nature of the demand forecasts, and we need to match the material flows across the facility with the information flows.

“We look as far as possible to ‘hybrid’ systems, combining manual and automated operations, to build flexibility into materials handling techniques as requirements change.

“This can often lead to a simpler automation solution. Many firms start by automating the most complex sequences thinking this will give them the best results. But if we start from a process view, looking at what the customer wants to achieve, we can apply the appropriate technology to each part of the value chain. The automation can be lower tech. We can use conventional technology as building blocks, scaling up to semi or full automation if required, without writing off the initial investment. Or take AGVs. In the early ‘80s there were a lot of AGV installations, often associated with Flexible Manufacturing Systems. Guided by wire embedded in a very high-spec floor, these were inherently complex and inflexible, and got a lot of bad press.

“Fast forward – we can take a standard vehicle, add on safety systems so it can work alongside manual systems, bolt on laser guidance and you have a highly flexible, cost effective solution which, because it is basically a standard piece of kit, is easy on service and maintenance. The next logical step is to use the same processes with the VNA range”.

Francesco Papale of Hyster concurs. “Take standard technology and fit a black box to it. Design your truck to be able to be part of automation, then if there are changes in the future – no problem.”

Papale adds that the traditional MHE and FLT manufacturers already have large and supportive dealer networks and remote diagnostics capabilities that reduce downtime, while “the truck can be operated manually if the unpredictable does happen”.

Dave Bull, sales manager for Dematic in the UK, maintains that, especially in e-commerce, “Everyone starts out with a fork truck, they don’t start with automation”. Part of his role, he says, is ‘hand holding’ as firms transition from the manual and thus inherently flexible processes they are comfortable with to pockets of automation and then full automation, while reassuring them that the automation they have invested in for a particular purpose will still be useful in a year’s time.

The approach, he says, is to “design a solution that looks like the future, but don’t put it all in on day one: design it all through then scale it back. That way, the customer has an assured growth path, and fast implementation when they eventually need it”. “But also, says Kuijpers of Witron, “We look at total door to door solutions influencing as much as possible of the supply chain beyond the four walls of the distribution centre to deliver the lowest overall cost. For example, we can do something smart in the DC that provides a strategy for store-friendly delivery, saving labour and creating overall cost efficiencies for customers. ‘Door to door’ matters in another way too. If you just have ‘islands’ of automation, you lose most of your money in the areas in between. So we drive our solutions from ERP – the customer may have multiple WMS/WCS systems across his DCs but for us these are just gateways – we integrate the solution at ERP level”.

Maxed out
The UK, Bull says, is indeed still at the ‘islands of automation’ stage, but this will change. “Automation isn’t about having less staff; it’s about doing more with the same staff. He cites ASOS, who have spent £45M on automation in the past three years. “Their busiest day was going to be 15,000 units/hour, with several thousand people pushing trolleys. Now they can do 35,000 an hour, but the same number of people are needed. You couldn’t have that level of throughput without automation. Most purely manual systems are maxed out – companies can’t grow. It isn’t just about finding the staff – where do you put the increased car parking and the bigger canteen? You can’t increase throughput by throwing the same tools at the same problems”, but at the same time, he acknowledges, “You can’t solve every problem with automation”.

Hence the logic of Dematic’s recent linking with Linde which in essence “allows the customer to know that they can access the capability and vision to increase productivity without throwing labour at it”.

Looking at the different automation technologies available, Bull says that “AGVs, laser rather than wire guided, are seeing a bit of a resurgence, but it’s got to be the right application. Shuttles are providing the main grunt in fully automated systems. There is, at least at the exhibitions, a lot going on with robotic picking, even picking individual items out of a box containing multiple skus, and often with the boxes presented by shuttle. And in the food industry we are seeing robots applied both to delayering and to multi-sku, multi-layer palletising”. He adds “For the future, robots will be cheaper, faster and more like a human being. So if you have a human operator at a station, you have to ask whether a machine can do this?”

Robots of course need not be tied to workstations, or even it seems to specific facilities. In a new twist on flexibility, DHL in the UK has bought four Sawyer robots to be deployed flexibly across 19 co-packing and production logistics centres in response to fluctuating demands. It is much easier to ship and reprogramme a robot than it is to hire and train additional temporary staff.

Denis Niezgoda is robotics accelerator lead at DHL, proving solutions for warehousing. “This is mainly in A-B movements, but also in face picking. We see robots working mainly in collaborative environments, moving around our workers, so there are a lot of positioning sensors and imaging systems to ensure safe movement”. Stationary robots performing repetitive work with high turnovers are already applicable and used, says Niezgoda, but order-picking robots will “follow workers around to support them” (or, since the robot knows where the next pick location is, they might actually be leading the workers around).
(Health & Safety, incidentally, isn’t just about stopping robots colliding with humans. Already in the US, a big driver for automation is the need to reduce liability for law suits covering back damage, repetitive strain injuries and similar).

But is complex automation oversold? Kuijpers says: “People are generally lazy – they see something and they want to have it. In the last couple of years, for example, shuttle systems have been the ‘golden egg’, but it is seen that they don’t cover everything perfectly. Companies say ‘We will always have the horsepower available’ but if you can’t use that power it doesn’t help you, it just wastes money. We have to look at specific situations and ask what type of automation will actually bring advantages”.

Automation is in relative terms far more affordable than hitherto, but Kuijpers says that “to keep the cost efficiencies, you have to get some standardisation, not just in materials handling equipment but in WMS configuration and so on. Many processes are independent of the nature of the business – there may be higher demands for safety or security, for example, but the processes of receiving, processing, despatching don’t differ much and can be based on standard elements.

“For example, pharma is about very small packages, and requires a totally different approach in physical handling to ‘ugly’ goods like furniture. Even within food retail, the Aldi approach is at pallet level and not touching the goods further. A Tesco, with many more skus, has less-than-pallet replenishment quantities. So for Aldi a manual or semi-automated solution can work without much loss of efficiency; but if you have 30,000 skus and a high dynamic of different requirements you waste a lot of time just driving around. We have to look at what level of automation and what concepts will have the digest effect on overall cost efficiency”.

Jungheinrich’s Richmond also sees a lot of interest in robotics, currently to deliver goods to manual stations but in the future the robotic picking of individual items. “But increasingly it is the software, not the hardware, that is the heartbeat of intralogistics solutions”.
The ability to reconfigure (and test) automation through the software rather than by ripping up the floor is a factor in what Richmond sees as a cultural shift making automated solutions more acceptable. “Change doesn’t put ‘business as usual’ at risk”.

Francesco Papale of Hyster-Yale Group says suggests that “cars are driving the new trend – every car has new automation features and this is driving change in the industrial sector. There is a new fleet of managers who welcome change and new solutions. Who notices cruise control on a car? So we can apply it to a warehouse truck – it’s almost expected.” DHL’s Niezgoda similarly sees a cultural change. “The next generation can’t imagine an environment without technology. We are working with robots that actively support workers but are not yet entirely autonomous as an initial step to familiarise the workforce – that’s why we have ‘humanoid’ robots with ‘eyes’, to create empathy – but that won’t be needed in a few years. Robots won’t replace people but we will see a shift in job descriptions as people orchestrate robots rather than doing most of the work themselves, although this transition may not be easy”.

Algorithms
At the moment, says Niezgoda, robots still do what they are programmed to do, but “we will see improvements in machine learning algorithms and improved visual recognition capabilities, providing more flexible and agile solutions”. (Hyster’s Papale also points to improvements in vision systems such as are being fitted to automobiles by companies such as Nvidia – “The ability to react to different situations will increase the flexibility of these vision based automation systems to emulate what operators are doing today”, he says. Bafani from Wipro also points to the need for ‘non-fogging’ vision systems for movements in and around chilled/cold stores).

“We are also working on making robotics intuitive enough for easy training,” Niezgoda continues. “There is still programming required, and a shortage of software engineers. We want to get to the point where, using for example voice recognition, you can simply tell the robot what to do next”.

Management and control techniques are developing rapidly with seriously affordable RFID and the Internet of Things. Richmond says that IoT is both fuelling and enabling the demand to know the position and status of every product in the supply chain and “the end to end information cycle will be the biggest driver behind the technologies”.

Dematic’s Dave Bull suggests though that the future “is not so much about the technology per se, it’s about having the knowledge and know-how to create new functions and techniques that can fast track goods through”.

With the convergence of hardware, software, and automation, Richmond says that firms like Jungheinrich “have to be intralogistics solutions providers – it is no longer adequate, and hasn’t been for many years, just to supply materials handling equipment”. Some competitors are addressing this by acquiring automation companies, although Richmond points out that Jungheinrich has been in automation since 1984 and has largely grown its skills and capabilities organically.

Conversely, Hyster, says Papale, prefers to collaborate with, rather than acquire, automation specialists – “we are the best in trucks, you in automation – we can share the solutions and service. (Otherwise we would have to create our own automation department…)”