Technology advancements in the current business landscape – and especially in the logistics industry – continue to afford and even demand that we innovate. This will create increasingly digitally enabled products and services that provide rich data and digital signals to convert historically “non-networked” physical products to a connected set of services and experiences enabled by web, mobile and IoT technologies. In this market, there is an opportunity to progress the value organizations deliver to clients from physical products and on-premises software to analytics derived from product usage as well as opportunities to directly impact and shape clients’ business outcomes. If the technology spectrum of advancement is a horizontal axis of progression, and the value organizations produce for clients an ascending axis – the intention must be to move portfolios “up and to the right.”
There are five steps to achieving this are:
- The creation of a software-defined logistics network;
- The development of operating systems for warehouses that improve processing through proactive notification of what to sort, what to pull and what to ship;
- A focus on global network expansion which is the synthesized use of cross-border networks and technology in collaboration with domestic assets to create differentiated service offerings, underpinned by;
- A focus on intelligent data services that leverage data science models and advanced analytics;
- A center of excellence focused on optimization and simulation capabilities (in the service of building digital twins of the physical networks and systems we operate) to improve service and spend performance
Software-Defined Logistics Network
Across industries, we have seen the Andreessen prediction that “software is eating the world” realized – as systems that historically relied on hardware and fixed configurations have migrated to become software-defined and software-managed. One great example of this is the software-defined wide-area network (SD-WAN) movement – where the wide-area networks that provide organizations’ internal networks have migrated from a collection of proprietary physical appliances to software-defined networks running on commodity cloud hardware platforms.
The benefits of this evolution are obvious – more agility, more flexibility and a greater ability to optimize networks in real-time based on signals derived from their operation. We see the same opportunities in logistics networks. In lieu of hard configurations of network routes and inflexible configurations of sort schemes, staging definitions, and freight load plans, these network elements must be moved to a software-definable configuration. Like the benefits seen in SD-WAN, as logistics networks become increasingly software-based, the implementation of optimizations of sortation plans and network routes derived from data science can be programmatically applied to the system.
Warehouse Operating Systems
As in a computer network, where the backbone of internet connectivity moves packets of data from one computing resource to another, software-defined logistics networks move physical parcels from one processing center to another – where the primary computing function on a parcel – sortation – is performed to route the parcel toward its destination using automation and robotics.
The benefits of these technologies are numerous – with the most obvious being that organizations convert what was historically a variable expense for each parcel into a fixed cost that is absorbed as they grow the volume in their networks. Organizations also see more consistent processing of the parcels, more complete data capture (including images, dimensions, weights, scans and movement events), and great precision in the path of the parcel through their facilities.
The automation of sortation also helps in creating a standard operating system for processing parcels, which can be designed to help the parcel sortation center prioritize what to process, when to process it, how to sort it, when to load it and how best to manage the outbound shipments to maximize transportation utilization while delivering on service commitments.
They can also be a collection of applications, IoT-powered in-facility experiences, directed calls to action via alerting frameworks, and sortation performance measurement and reporting that helps ensure that parcels experience minimal latency and make the required critical cut times for outbound shipment from the facilities.
Global Network Expansion
There are a variety of logistics services made available by organizations across the industry, including standard U.S. domestic delivery, parcel returns service, expedited parcel delivery and cross-border parcel delivery. Today, there is an increasing number of opportunities to combine these services in logical ways to create new and compelling service propositions. For example, organizations can use a combination of first mile providers, cross-border capabilities, and U.S. domestic delivery networks to create compelling new logistics service lanes from a foreign country to U.S. consumers.
Similarly, organizations can take existing domestic service technologies and apply them in other countries (with different first and last mile partner solutions) to offer new domestic standard delivery service options in foreign markets. Finally, through data science, organizations can optimize the way in which they render the underlying physical processes that assure expedited parcel delivery to improve the cost basis of those services. Using a combination of software-defined carrier integrations and network systems, as well as API-based service capabilities, organizations can rapidly synthesize new offerings for the global logistics market.
Intelligent Data Services
There are significant opportunities to apply data science to forecasting, sortation planning, network simulation and optimization and anomaly detection. Focus must next be placed on improving and unifying the forecast process used across operating processes. There are new machine learning-based models to predict origin forecast, by site – which tie client demand forecast to operating centers.
Similarly, tooling and processes to allow exigent factors like client promotional sales, new client launches, weather events, and other factors that fall outside the historical learning of our models are being built across the industry to be incorporated into the complete operational forecast. This singular source of forecast can be used to build consistent and coherent network-wide estimations of final sort densities, truck utilization, truck schedules, and labor plans for facilities.
Similarly, it is crucial to constantly improve intelligent data services for trade classification, parcel classification (is this a bag or a box?), and the detection of anomalous parcels (which fall outside expected temporal, spatial, or process norms). By centralizing the development and curation of these data services, and putting them under a common API framework, we create a unified data intelligence layer across applications, processes, and analytics.
Simulation and Optimization of Digital Twins
One of the most fundamental practices of engineering is the ability to use a set of pragmatic assumptions about a physical system to create a mathematically tractable problem to design a working solution for the physical world. This approach has been used for centuries, and everything from the beams that support bridges to the thermodynamic processes that underpin power plants are designed and sized using these methodologies.
As technology advances, increasingly dynamic and complex systems can be modeled using software tools and methodologies. In large, complex systems like delivery networks, organizations must create digital representations of these physical systems (sometimes referred to as digital twins) to allow them to simulate network performance and “search” for alternative network and processing scenarios that better optimize the use of resources (and thus cost) to deliver the required service performance.
Today, organizations must create competencies in each of these areas and demonstrate the ability to perform analysis, make network changes, and realize double-digit service performances – all in a span of weeks.
It is this combined set of capabilities – the software-defined logistics network, warehouse operating systems, and organizations’ abilities to create global logistics solutions – and underpinning these with both intelligent data and simulation and optimization capabilities – that provides them with the ability to design delivery, design returns, or design cross-border services that fit their clients’ unique needs.
The path forward is clear: Embedding data science, software-enabling systems, and engineering networks to perform for clients using all these capabilities will directly create better consumer experiences and better outcomes for clients.
Dr. James Fairweather is the executive vice president and chief innovation officer at Pitney Bowes.
If you enjoyed this article and want to receive more valuable industry content like this, click here to sign up for our digital newsletters!
Leave a Reply