Picking is a crucial link in retail supply chains, and the pressurised environment of e-commerce is creating some dazzling efficiencies. Johanna Parsons puts on her sunglasses.
The success of e-commerce is entirely reliant on the successful supply chains behind each online transaction. The link in those chains that probably comes under the most pressure is the pick face, the point where the flow of goods intersects with the purchase order and begins its journey to the end user. But pressure makes diamonds, and the pressures on order picking are forging some valuable results.
As with the formation of diamonds, the process of acquiring a market beating picking operation is also complex, rather time consuming and often heated. Many businesses first have to face the uncomfortable fact that extant systems are simply not up to the standards set by the omni-channel market. “The growth of e-commerce and omni-channel retailing has reached meteoric levels in the last five years, with the result that many processes and legacy systems are now inadequate,” says Eric Carter of Indigo Software.
Steve Richmond, director – logistics systems Jungheinrich UK says the pressure is putting logistics in the spotlight. “E-commerce is having a significant impact on organisations and how they think about the logistics supply chain in its entirety – particularly within intralogistics…
“For organisations dealing in commodity products in particular, supply chain processes must be robust, transparent, and deliver a high degree of accuracy and traceability for customers to be satisfied with their order. If they don’t, it’s all too easy for consumers to switch to another retailer who will offer the same product at the same price, but with a better level of service,” says Richmond.
The problem is not a lack of investment but rather where money has been directed. Carter says “Lots of businesses have made significant investments in their front end websites but haven’t focused too much on the warehouse. Their processes are creaking at the seams and they need a system that can back up the impressive growth rates these businesses are seeing. Smaller companies in particular are investing in WMS technology, to help support their e-commerce activities,” he says.
The merits of warehouse control IT are now widely appreciated across the industry, and many software developers are providing a plethora of more sophisticated offerings. Though Alex Mills of Chess Logistics Technology and ProSku points out that – especially for the budget-conscious operator – it is perhaps best to keep it simple.
“There is a lot of talk about Blockchain and similar technologies in the logistics sector at the moment. Although this could have some utility in the longer term, for the time being companies can make better use of their time and money by focusing on improving the performance of their warehouse operations to simplify processes, remove wasteful tasks and increase data accuracy – all of which is supported by a properly designed and implemented WMS,” says Mills.
Richmond agrees that IT is of increasing importance to operators, but he says the most impressive results for throughput are yielded through highly automated goods-to-man picking processes. “Here, the back end of the material flow is fully automated and will be bringing products automatically to pick and pack stations, generating very high throughput, accuracy and, most of all, efficiency throughout the supply chain.”
Mike Hilton of TGW Logistics says that the escalating demands of e-commerce mean that such high-end automation is becoming more justifiable for more businesses. “In terms of end user expectations, speed of delivery and in-store stock availability remain key. Both are reliant on a fast and accurate picking process, which is why interest in and the adoption of reliable and dexterous automation for picking operations continues to grow.”
But all-out automation is not the only route to faster fulfilment. There are a myriad of technologies that make the human picker a faster conduit to the data at their finger tips. A recent example is Honeywell’s new wearable computer that it says improves typical transaction times by five seconds per transaction.
The 8680i Wearable Mini-Mobile is worn on one hand and provides a two-button interface and a display of workflow instructions for picking, sorting, put-away and packing via Wi-fi and API. The hands-free cordless device eliminates the need to pick up a handheld scanner, scan the item and then set the scanner down again which in scan-intensive environments translates to a big saving.
Richmond explains that technology like this shows how different approaches can thrive alongside each other. He says there is still a role for less flashy picking systems. Every warehouse, and indeed every part of the warehouse has different requirements.
“Automation doesn’t mean that every aspect of the supply chain has to be automated – just the parts of it that will benefit. Hybrid systems are therefore increasingly attractive, whereby a combination of manual, semi and even fully automated systems are optimised by an over-arching warehouse management system to optimise material flow. In a lot of cases, this requires less capital outlay and delivers a quicker ROI than if the organisation transitioned straight to a fully automated system,” says Richmond.
Less capital outlay and a quick return on investment are key requirements for most operators, particularly those making their first major investments in intralogistics. Mike Alibone of SSI Schaefer says there are certainly options for firms with a more modest budget.
“It all depends upon the size of the system and what they requirements are. Beyond renting/leasing a system it may be difficult to cut costs on implementing a large scale operating system. Aside from the fixed IT control system cost, however, starting with a small, modular system and adding to it as the business grows is clearly an option for many but not all.”
He gives the example of the SSI Flexi Shuttle, a modular concept which combines shuttles, lifts, and storage of variously sized conical and cubic bins, cartons and trays with automated storage, buffering and sequencing in one racking system. And the firm’s Weasel automated guided vehicles have drawn attention for their low cost.
Edward Hutchison, managing director of BITO advocates taking a stepped transition to automated picking. “A good example being a combination of ‘middle-ground mechanisation’ and intelligent software that provides a first step on the automation ladder, while at the same time giving scalability and agility. This, of course, is vitally important in the fast evolving home delivery sector.”
He gives an example of a project BITO has developed with an automation supplier that provides ‘put wall’ picking systems to fulfil home delivery orders. This involves picking several orders at once and taking them to a “put wall” which is a bay of shelving that allows the picker to consolidate the items into containers in pigeonhole positions for each order. “Once an order is complete the light switches on to inform the packer on the other side of the shelving that it’s ready and waiting. The packer can then pull the container holding the complete order through for it to be packed as one delivery,” says Hutchison.
Swisslog head of sales (UK), Shane Faulkner says that delivery speed is not the only area where expectations have risen. “Other added value activities such as gift-wrapping and personalised messages are now commonplace. This is where cobots come in – robots that work collaboratively with humans – as robots today are seen less as a substitute, and more as a supportive and assisting medium for human workers,” says Faulkner.
For a long time robot collaboration simply meant an ASRS delivering totes to a human picker. But the intelligence of these learning robots means that the actual pick is set to become more sophisticated and automated.
Faulkner explains that Swisslog’s AutoPiQ robot picks the items that it is able to pick, which can be up to 95 per cent of the warehouse’s product range, before a worker finishes the order at the pack stage. AutoPiQ uses a smart portable robot cell, ItemPiQ, which uses an intelligent vision system to detect unknown items and with four different gripping possibilities it learns how to handle even fragile products. Faulkner says it is this machine-learning concept that opens up opportunities in logistics.
“With a human worker involved in the process, each individual order can still meet end user expectations in personalisation. In manual warehouses, this would add pressure on workers and increase the risk of errors,” says Faulkner.
A new collaboration between Tompkins Robotics, and RightHand Robotics is focusing on the blend of intelligent learning and manual handling. The project uses t-Sort which performs much like a conventional tilt tray or cross-belt sorter but with independent robots, allowing any-to-any flow that is easy to reconfigure as business scales up or to accommodate seasonal peaks. This is linked with RightPick which picks individual items, handling hundreds of thousands of different items using machine learning with an intelligent gripper that works alongside industry-leading robotic arms.
TGW’s Hilton agrees that it is inevitable that robots will muscle in on the physical pick action itself. “All the major integrators are actively investing in developing robots that can solve such problems as emulating the human hand, as well as intelligently recognising/learning how to not only handle different shaped products but also to identify how to effectively fill a tote or carton with these products.
“This is being combined with the ability to react if the wrong quantity of products is picked and to sense/rectify if the product is dropped outside of the target.” He describes how TGW’s “FlasPick” uses a self-learning robot “Revolution”for goods-to-person smart piece picking . The benefits are obvious in that intelligent self-learning robots like TGW’s Rovolution can work every hour of the day, every day of the year,” says Hilton.
Labour saving is the big selling point of the robot, and labour is now a precious commodity. The supply of warehouse workers is getting more expensive and less reliable, whereas demand for labour is rising, particularly with the extreme peaks of the retail calendar.
“Picking is often considered one of the most labour and cost-intensive processes in intralogistics,” says Swisslog’s Faulkner. “It takes valuable time for individual order lines to become a complete order, and with tasks such as finding the correct items, ensuring they are counted, separated, checked and labelled; as well as documenting each pick using a scanner or touchscreen monitor, it’s clear that picking is much more than just grabbing something out of a bin.
“Shared picking principles are already more commonplace to accommodate this and, among other things, order picking robots offer clear benefits when it comes to reliability, as they do not require breaks or sick days. Switching from a two-shift operation to a three-shift one is possible at any time too, allowing companies to respond flexibly and efficiently to order peaks such as the holiday season,” says Faulkner.
The latest reports from a trial by GEODIS underline just how effective this approach can be. The pilot of collaborative robot picking with an online women’s apparel client used 30 autonomous mobile robots from Locus Robotics in a 139,000 sq ft warehouse in Indianapolis, USA. 80 per cent of the units were picked to the robots daily. GEODIS says employee productivity doubled and there was at least 50 per cent reduction in time to train new employees.
Marie-Christine Lombard, GEODIS chief executive officer says: “The labour market is tight, especially during peak seasons, and we want to enable our team to better execute for our customers. And in this case, the technological support of robots effectively solved the challenge.”
The efficiencies promised by robotics are seductive, but they haven’t quite taken over yet. Indigo’s Carter has reservations about how widely this approach will be taken up.
“Robotics and stock to person picking methods are revolutionising picking processes in the warehouses that have made these investments to date, but it’s worth putting this into perspective. For every five warehouses that have introduced robots, there are 150 that haven’t as yet,” says Carter.
He says that there is no one size fits all system, and many retailers are adopting multiple picking systems to get the balance right. “Companies are increasingly implementing mixed mode picking methods to get the highest possible efficiency and flexibility, for example, implementing a combination of RFID, voice technology and mobile devices – it’s a mix and match of technology to meet the needs of a business,” says Carter.
So whether it’s intelligent data flows or simply wise investments, there are undoubtedly some gleaming efficiencies to be gained from optimising the picking process.
And from wearable picking devices, put-walls and WMS systems to learning robots and a connected warehouse, there are seemingly more ways than ever before to polish a picking system into something that shines.
Case study: French retailer takes KNAPP’s E-Grocer system
AuchanDirect.fr – a subsidiary of French food retailer Auchan, is delivering an average of 2,800 online grocery orders per day from a new distribution centre equipped with automation from KNAPP.
The retailer founded a new 136,000 sq ft DC in Chilly Mazarin, south of Paris using KNAPP’s E-Grocer automated systems to get online orders to customers’ homes within 24 hours.
KNAPP’s E-Grocer system is made up of the OSR Shuttle solution for storage, picking and dispatch buffering; ergonomic Pick-it- Easy workstations for order fulfilment; and automatic stacking machines. Products are picked directly into bags suspended within totes, each one taking three bags of different colours for the various food types. Frozen products are picked separately into special cool boxes.
Incoming goods are decanted at one of 11 stations and then conveyed to the appropriate warehouse area. Fast moving items are picked manually from pallets and flow racking into totes on a conveyor system, with the help of hand-held RF devices. They are then grouped into batch orders to streamline the path through the warehouse and reduce picking times.
Medium and slow moving items are stored in the OSR Shuttle system which supplies source totes in the correct sequence for order picking at the 12 Pick-it-Easy workstations. The operator is guided through each order with the help of an intuitive user interface. Fruit and vegetables are picked directly from shelving into the bags of totes transported on a trolley.
The main OSR Shuttle features 34,816 tote locations over 16 levels and there is a separate OSR Shuttle store for fresh goods with 10,880 locations, plus another OSR Shuttle for ambient goods with 4,096 locations. The latter is also used as a dispatch buffer. As soon as the orders for a particular route are completed, the totes are automatically retrieved from the OSR Shuttle in the fresh food area and/or from the dispatch buffer OSR Shuttle store and automatically stacked in reverse delivery sequence.
This article first appeared in Logistics Manager, June 2018