20th anniversary features: Supply chain IT
“I have seen the future – and it works,” observed American social commentator Upton Sinclair almost 100 years ago, after a visit to post-revolutionary Russia. At various points over the past 20 years, similar sentiments have echoed repeatedly through the pages of Automotive Logistics, as successive technologies have redefined the art of the possible.
Sinclair’s enthusiasm about communist Russia, as we now know, was misplaced. And likewise, in the world’s automotive supply chains, there has been a more sober reassessment of the real impact of many of these supposedly revolutionary advances.
That isn’t to say that the technologies themselves don’t work, nor is it to say that those same technologies haven’t delivered genuine supply chain transformations in other industries.
For proof, simply order something from e-commerce giants Amazon or eBay, whose respective launches only narrowly pre-date that of Automotive Logistics. A succession of impressive supply chain technologies will promptly deliver the ordered item to your door, along with a track-and-trace capability that lets you track progress every step of the way.
In the world of automotive logistics, however, aspirations are of a different nature. Having invested over the past 20 years in data acquisition technology, the industry is now casting around for ways to make sense of that data.
“The world of logistics is generating more data than ever before – telematics data, GPS data, data from hand-held devices, photographic data, and RFID and barcode data,” points out Paul Nurse, chief executive of ProAct International, a provider of logistics management and ‘control tower’ solutions. “What it isn’t doing is getting significantly better at taking all that data and making sense of it: we know where something is, but should it be there? Is it early? Is it late? Is it even wanted? And where should it be going next? The industry has the data, but not necessarily the interpretive capabilities to match.”
A key difficulty in acquiring that interpretive capability is that data generally moves too slowly, says John Miller-Wilson, director of vehicle logistics at Sovereign Business Integration Group, provider of a software-as-a-service-based vehicle logistics management solution originally developed for Ford and today used by Suzuki in the UK.
“FTP servers and flat-file data file transfers are still the bedrock [on which] information moves along the industry’s supply chains. Despite all the advances that have been made – think of hand-held devices, RFID and ePOD [electronic proof of delivery] technology, for instance – the underlying paradigm is often batch files and once-a-day file transfers,” he notes. “Consequently, everything works in batch mode – and when the injection of real-time data is attempted, it generally causes problems, as conflicts arise with the central ‘batch-driven’ view of the supply chain.”
Gary Allen, vice-president of supply chain excellence at US-based 3PL and logistics specialist Ryder System, meanwhile, highlights the differences between the automotive industry’s broader technology aspirations and the technology extent in its current supply chains. One of the big jumps that the industry has yet to make is a move away from legacy processes like electronic data exchange (EDI) and advance shipping notice (ASN), which typically have a lag in transferring information, to more instantaneous systems like application program interface (API).
“The industry is up to date in its products and in its commitment to technology standards but in terms of supply chain visibility, it is falling behind,” Allen observes. “Walk through an automotive plant and there are still a lot of dated green screens to be seen: automotive companies have been investing in the technology in their products, but less so in the visibility and control technology in their supply chains. The foundation is still legacy technology, planning in terms of ASNs rather than APIs, batch data rather than real-time data, and a reliance on EDI and fixed schedules.”
Still a one-dimensional process
Yet if the diagnosis is clear, the remedy remains murky. The fact is that batch-based data processing is not only cheap and simple but also reflects many aspects of how the industry works. Trains, ocean-going car carriers, even road transporters – conceptually, these are all ‘batches’ of vehicles. Likewise, on the inbound side, EDI-based transactions and sequences reflect the industry’s processes like a glove. Not least, of course, because that’s what they were designed to do by the industry’s own representatives on bodies such as the Automotive Industry Action Group (AIAG) in North America and Odette in Europe.
“Five years ago, people were saying that EDI had no more than five to ten years left – yet 90% of the world’s largest manufacturers are still using it,” observed Steve Keifer, then vice-president of industry and product marketing at EDI provider GXS, in an interview with Automotive Logistics back in 2006. “Analysts are now saying that it’s going to be around for another 20 years at least, while we’re regarding it as ‘here to stay’.”
Moreover, the industry has proved adept at finding ways to sidestep some of the new technologies supposedly coming to its rescue. RFID, for instance, was the ‘next big thing’ 20 years ago, yet has still to make a substantial mark on the industry. In the same interview more than a decade ago, Keifer also noted that XML-based messaging – popularly adopted in other industries – was struggling to displace EDI. Instead, manufacturers were increasingly transmitting EDI over the internet.
“The world of logistics is generating more data than ever before – telematics data, GPS data, data from hand-held devices, photographic data, and RFID and barcode data. What it isn’t doing is getting significantly better at taking all that data and making sense of it: we know where something is, but should it be there? Is it early? Is it late? Is it even wanted? And where should it be going next? The industry has the data, but not necessarily the interpretive capabilities to match.” – Paul Nurse, ProAct International
The humble linear, one-dimensional barcode has also proved somewhat difficult to dislodge, despite having been invented as far back as 1949 and first deployed in a supply chain application in June 1974, when a checkout clerk passed a ten-wafer pack of Wrigley’s Juicy Fruit chewing gum over a barcode scanner at a supermarket in Troy, Ohio. Its persistence is even more surprising given the invention of the two-dimensional QR barcodes by Toyota subsidiary Denso back in the early 1990s, specifically for automotive industry supply chain applications. The objective: a 2D labelling technology that could handle more data than traditional 1D barcodes and would be better suited to damage-prone industrial environments.
“Whereas 1D barcodes can only carry around 20 characters of information, 2D barcodes carry thousands of characters,” explains Andrew Moses, managing director of The Config Team, a specialist SAP supply chain consultancy and implementation provider. “Labels can carry information such as batch code, production date, plant of manufacture, expiry date and production line. What’s more, they can be scanned from acute angles and are in less danger of losing information because of scratching or damage.”
And ironically, given RFID’s much-touted ability to be read at a distance, 2D barcodes turned out to have useful distance-reading capabilities, too.
“Whereas 1D is read at a range of 20-30 centimetres, 2D can be read at distances of 1.2-2.4 metres,” says Ronen Dahari, head of research at barcoding and auto-ID specialist Zetes Industries. “And with wide-field optics, you can read a hundred barcodes in less than a second, even at that distance.”
Yet, like batch-exchanged flatfiles and EDI, 1D linear barcodes remain broadly in use within the industry, partly due to its innate conservatism, partly due to ‘network effects’ – the need for a lowest common denominator lingua franca that can be read by everyone – and partly because of the lack of an investment case.
“Standardisation for the sake of standardisation doesn’t add value,” was how Kevin Mixer, then automotive research director at analysts AMR Research, explained the persistence of 1D barcodes in an Automotive Logistics article back in 2004. “If today’s barcodes are working well, does changing the format reduce costs or provide greater efficiency? And the answer is no, it doesn’t. So it’s unlikely to happen.”
From seeing data to interpreting it
Nevertheless, an insistence on a sound investment case before investing in new technology is not the same as a flat refusal to make progress. The industry might be cautious with its investment dollars – especially after enduring the 2008-2009 recession – but it certainly isn’t set on standing still. Put another way, there’s a firm improvement agenda, and what is required is affordable technology platforms that can meaningfully help to deliver on that agenda.
What might those be? This, say many of those close to the industry’s technology platforms, is really the wrong question – as past experience with technology like RFID highlights only too well. Instead, goes the argument, a better approach is to remain focused on the industry’s main improvement drivers, and then pick out broad trends in terms of the likely direction towards delivering those improvements. In other words, in terms of figuring out what the next 20 years might hold, the best answer is likely to be a direction of travel, rather than a destination.
Improved decision-making, for example, ticks a lot of boxes. From a technology point of view, says Dominic Regan, senior European director for logistics applications at enterprise applications firm Oracle, this calls for the focus to move away from generating data to processing that data and getting value from it. In other words, if the past 20 years has been about the technologies that create and generate data, the next 20 years will be about making better use of that data.
“The question is no longer: ‘Where is it?’” he explains. “Instead, it’s about interpreting the deluge of data that is arriving. We know where something is – what we don’t know are the implications of it being in that location.”
Achieving that understanding, he stresses, calls for more than individual so-called ‘point solutions’ such as analytics, advanced scheduling, or visualisation. Not least, he adds, because an answer is only truly of value if it is actionable. As yet, the industry still has too many organisational silos and demarcation lines for situational changes in one part of the end-to-end supply chain to be reflected in others.
“If a ship arrives in port a day early, what does that mean for the assembly sequence and the inbound logistics flows? In theory, there may be opportunities – in practice, those are questions that aren’t being asked.”
And to ask them, the industry will need both better interpretive capabilities and better integration to tie all these disparate data sources together within a single ‘control tower’. Among carmakers, BMW looks to be moving in such a direction of travel, both from a risk management point of view, and an operational management perspective – the latter seen most recently with a real-time shipment and inventory tracking system rolled out this summer at the company’s Spartanburg plant in Greer, South Carolina, using logistics integration specialists Leogistics as the system integrator.
“The future lies in integration,” stresses Leogistics chief executive André Kaber. “We bring together information, integrate it with other information, and then apply visualisation techniques and business rules – so that if an incoming trailer is going to be late, that information updates all the other systems that need to know this and take action on it. Map data, weather data, GPS data, sensor data, telematics data: information is always more powerful when integrated with other information.”
Building blocks of the future
Such integration and interpretive agendas will in turn help to build the case for emerging ‘point solution’ technologies, reckon those close to the automotive industry’s IT and logistics platforms. On their own, these might struggle to make a case for adoption – but if they help to deliver a meaningful increase in the industry’s predictive and interpretive abilities, barriers against adoption could soon fall.
ProAct International’s Paul Nurse, for instance, points to the disruptive power of Bluetooth tags, as opposed to RFID technology. “RFID tags aren’t inclusive enough: specialist equipment is needed to read them,” he notes. “Stick a Bluetooth tag on something and anyone in the supply chain with a smartphone can scan it.”
For his part, industry observer Matthias Berlit, vice-president of manufacturing logistics at logistics optimisation provider Inform, highlights the potential of blockchain technology as another example of a potential direction of travel.
“Blockchain’s ‘distributed ledgers’ are an enabler of integration and visibility,” he points out. “Blockchain data is not only visible to everyone in the supply chain – and so acts as a single integration layer – but also provides a layer of trust.”
Pilot logistics applications of blockchain technology are already emerging, he adds. One such, jointly built by logistics technology provider Marine Transport International and Agility Sciences in collaboration with scientists at the University of Copenhagen, has connected a supplier, shipper, load point, customs and terminal on a shared blockchain ledger, with each party benefitting from automated data flows, enhanced security and greater resilience.
“The business case for connecting supply chains using blockchain is very strong,” says Jody Cleworth, chief executive of Marine Transport International. “The interface is easily adaptable to existing systems, so there is a very low barrier to entry. And a blockchain-enabled supply chain is highly resilient to cyberattack – a copy of the data is stored on each node on a decentralised network, meaning that even if one node is compromised, the data is safe.”
If advances in real-time visibility and data exchange in the automotive supply chain were slower than many might have hoped over the past 20 years, the next few years, in which swifter data connections across sources and analytics look set to take hold, should make up for lost time.
“RFID tags aren’t inclusive enough: specialist equipment is needed to read them. Stick a Bluetooth tag on something, and anyone in the supply chain with a smartphone can scan it.” – Paul Nurse, ProAct International