not on facebook


A postal network for the digital century

11 Dec 2018, 18:48 ... ry/198123/

The world’s postal and delivery services have been undergoing profound change over the past few years, whether that’s having to scale up to meet increased demand or expanding and modernising their networks.

The business to consumer market is growing rapidly and parcel volumes have jumped on the back of a surge in e-commerce; customers are also looking for multiple delivery options which has led, in turn, to further investment.

Today, consumers have more choices and options when it comes to delivery services, and to be successful postal and parcel companies are having to focus on delivering a last mile service that keeps customers in control of how, when and where their parcels are delivered.

Technology has a crucial role to play and Lockheed Martin is working with postal operators around the world to provide innovative and cost-effective technologies to support and help improve postal networks.

Lockheed Martin is a systems integrator of postal technology and services and, according to the company, takes a holistic approach to the industry not only addressing technical but contractual and commercial issues as well.

Its technology is currently being used to sort around a quarter of the world’s mail through the use of its address recognition systems. It is, however, a market that’s been in decline for some time – due to the rise of email.

"Traditional letter sending has been falling for many years," explains Emlyn Taylor, Managing Director of Lockheed Martin UK System Solutions. "However, the collaborative approach taken by companies like Lockheed Martin and Siemens, with the postal operators, means that automation has grown dramatically, and that’s helped ensure that efficiencies have off-set that decline. As a result, most postal authorities are still able to operate profitably.

"State-owned and independent postal operators face domestic and international challenges when it comes to their supply chains and it’s not just falling mail volumes and revenues that’s the problem," says Taylor. "E-commerce volumes and more consumer choice are both having a considerable impact on the industry."

In response Lockheed Martin is investing in new capabilities in terms of address recognition, data management and systems integration and using advanced image analytics technology and artificial intelligence to better organise and distribute the post.

The company’s specialist postal automation technology - MEARS (Mailpiece Events and Attributes Recognition System) - which was first put into operation with UK customers back in 2010, has now been future-proofed until 2023, and should help keep UK postal sortation technologies ahead of the competition in the growing e-commerce market – where the delivery of packages and parcels has soared on the back of platforms like Amazon and e-Bay.

"While automation has certainly benefitted the efficient sorting and delivery of the mail, it’s far harder to use automation to help with e-commerce," Taylor points out. "When it comes to parcels, the issues centre on how can current systems and techniques cope with the variety of labels and formats that are being used?

"There are other issues too, such as how can customs payments in multiple currencies be automated and how can automation be deployed when it comes to VAT collection, multiple currency declarations or when looking to counter fraud?"

Up until relatively recently those issues have tended to be ignored or if a solution had been developed and deployed it was found to have performed poorly, Taylor suggests.

"Rather than use automation or technology parcel delivery has tended to remain reliant on manual sorting – the available solutions either don’t work or aren’t trusted."

According to Taylor e-commerce, raises serious questions about how the industry can remain profitable knowing that the only practical solution has been an increase in the use of manual labour to sort packages.

"Packet and parcel volumes are increasing and there are particular problems for postal services especially at peak times.

"Flexible collection times and new delivery services add to the complexity of the market.

"Our response has been to bring greater agility to the operation and integration of new and legacy systems."

Technology to the rescue

Lockheed Martin has been working with the postal sector since the mid-1980s when it started to introduce automation into the postal service.

"Then it was an industry that employed 100,000s," says Taylor. "When it came to automation our role was to enable the post to be moved around quickly and enable humans to better interact with devices and machinery."

The size of the US Postal Service (USPS) meant that it took a leading role in the development and deployment of technology.

"Working with USPS Lockheed deployed optical recognition technology for the first time to sort the mail without requiring the size of workforce that hitherto had been needed – we were deploying rudimentary computers and artificial intelligence.

"Over a twenty-year period, sorting machines became more accurate, better able to match characters against a data set as well as able to work with other manufacturers’ software.

"When it first started optical character recognition (OCR) was at most 5 percent accurate. By the early 2000s it was 99 percent.

"That technological advance was crucial to a market like the US, where the mis-direction of post can prove extremely costly."

Postal services are relatively easy to automate, according to Graham Urqhuart, a Computer Systems Architect with Lockheed Martin.

"The challenge lies with their efficient management.

"That’s not the case when it comes to parcels. As Emlyn (Taylor) has already explained, postal operators across Europe are seeing a dramatic increase in volumes much of which comprises of small, light items, with low-value contents, mostly from the Far East.

"The complexity of the labels being used means that they are very difficult to read using conventional OCR technology," Urqhuart explains.

"They are too complicated and busy for the software to identify the correct information to read: low read rates and high errors give poor machine efficiency, slow sorting and significantly increased costs due to higher manual sortation rates."

But that’s not the only issue for the postal service. There is a growing requirement, especially in Europe, for postal operators to capture details such as content, weights and values, as well as the recovery of VAT for the Customs authorities.

By 2020, in Europe, full customs and VAT payments will be needed to be paid on all parcels. That will be a massive problem for the region’s national postal services, according to Urqhuart.

"The only way to do any of this, at present, is through the manual extraction of data. This is slow and can seriously impact sortation rates – leading to delays and increased costs. There is, currently, no standard way of addressing parcels and parcel sizes remain a challenge.

"Under the terms of their contracts with governments, national postal services are under a legal obligation to deliver parcels, whatever their size or shape. If they receive a fish with a stamp on it, they’re obliged to deliver it!

"That’s not the case for couriers. They can use a standardised process and use bar codes which makes it easier to automate the process.

"Amazon, for example, has a standard packaging format. National postal networks have to handle millions of variations in parcel formats, and they work with a universal service obligation which means they have to sort everything if it has been paid for."

An automated real time capture of information from labels is therefore essential if European postal services are to address this as well as meet more demanding customer expectations.

Minerva is a deep learning machine technology that's already being deployed with postal operators


In response, Lockheed Martin has developed Minerva, a machine-learning based solution that’s intended to improve the sortation and extraction of information from complex cross-border ecommerce packet labels.

"Minerva is a deep machine learning technology built on knowledge that we’ve developed over many years. We are in a market where profits are harder to come by, so we need to work closely with the supply chain to come up with innovative solutions.

"As a company we spend a lot on research and development and when we looked at the postal market a few years ago, the transformation we saw coming in the parcels market meant we needed to look at how we could help our partners to address that.

"Minerva is already in operation with PostNord Sweden, where we have demonstrated that it can significantly speed up automated sorting of cross border packets and parcels," explains Taylor.

"It provides real time identification of delivery addresses; can extract Customs information for the recovery of VAT and, crucially, it can be integrated with existing sortation systems."

The company is looking to deploy the platform in the UK early in 2019.

The upgrade to PostNord’s sorting technology also includes a Next Generation Sortation Platform (NGSP), a flexible business-rule driven workflow engine, which brings increased flexibility to PostNord’s operations throughout Sweden.

"Minerva makes use of sophisticated artificial intelligence routines to provide greatly improved recognition performance for e-commerce packets and helps to increase the ability to process a complex and rapidly changing mail stream," says Taylor.

Technology, like Minerva, also has an important role to play when it comes to security.

"The postal service has always been in the frontline when it comes to security but today, enabled by technology, we can implement more layers of security.

"Data rich platforms, like Minerva, provide an opportunity to collect data. It can spot trends or patterns which could be suspicious and provides a platform that can bring together multiple forms of data to form one essential piece of intelligence."

Deutsche Post DHL’s Trend Research team has identified several trends that could impact the postal and logistics industry over the coming 5-10 years and according to its findings, published earlier this year, innovation will be more important than ever, as digitisation continues to transform the industry.

According to DHL, customer centricity will be key to providing a faster and more convenient logistics experience – and a key focus for future innovation will come in the last mile which could see the integration of logistics services into smarter home environments.

Technology will become widespread as the Internet of Things and Artificial Intelligence grow in importance and the spread of next-generation wireless networks will increase the value derived from connectivity in the wider supply chain.

Highly repetitive, physically intensive tasks will be aided by technology, enabling people to do more meaningful tasks that require management, analysis and innovation.

"That could see the use of robots," suggests Taylor. "Looking at the postal service, letters are now fully automated. We’re starting to deploy platforms capable of doing the same with parcels.

"Our aim is to deliver the ‘art of the possible’ and squeeze out additional profits for the industry and that means, delivering a mail centre or warehouse in which there is as little human intervention as possible – every human touch is a cost to the business."

Robotics, autonomous vehicles, artificial intelligence all have a role to play in the postal services of the next few years.

While the deployment of these technologies will vary from country to country the nature of the postal service is set to change profoundly and permanently.

Previous page Next page

Page 1 of 1