Software solutions that orchestrate connections between processes and close the gap between visibility and action offer the most promise to deliver supply chain transformation. aThingz is leading the way in retooling the logistics planning and execution space.
Introduction
“Transformation” is an overused word in business today, particularly as it relates to supply chains and supply chain management. Seemingly everyone is seeking a “transformation,” in which the supply chain moves from one operating paradigm to a new one, with a higher level of competitiveness as reflected by customer service, cost, profitability, and return on investment.
But are such transformations happening? And, how do you know that you have moved from one operating paradigm to another?
Certainly, in the past fifteen years, supply chains have been transformed into consumer-driven engines. The poster child for this is Amazon. Customers have come to expect myriad choices in SKU counts, products, and product variants, delivered to them with increasingly precise delivery methods within increasingly precise delivery windows. And this is not just about B2C supply chains. Downstream expectations of consumers are quickly driven upstream to B2B relationships, tightening the screws on everything from inventory to production schedules to delivery windows.
With the advantage of twenty years of hindsight, certainly much has changed, and much has been learned and achieved.
However, those who have been working in supply chain management and software for the past few decades know that much more can be done. A behind-the-scenes view of how the sausage is made is often not pretty. This was on full display to the entire world during the pandemic. Lora Cecere exposed a little bit about how the sausage is made in her post last month (“Please do not AI this”). In her article, Lora includes a common supply chain management diagram that purports to represent “end-to-end” planning. As she rightly points out, it does not.
Here we discuss a fundamentally different approach in the context of logistics planning and execution software. This approach is driven by a company called aThingz. aThingz has used first principles thinking to upend current approaches and develop a software product that has delivered hundreds of millions of dollars of value to one of the largest manufacturers in the world.
Logistics Software Landscape
A typical modern logistics software environment consists of visibility platforms, distribution planning software (also called replenishment planning), transportation planning software as part of a TMS, and financial planning software that is part of the finance department. Surrounding this software is order management, inventory optimization, plant scheduling that may be part of an MES, warehouse dock scheduling that may be part of a WMS, and material requirements planning software that may be part of an ERP system from SAP, Oracle, or elsewhere. Furthermore, many companies drop various analytics systems in on top of these functional systems to collate data and provide performance management and reporting on such things as cost-to-serve and utilization.
To add fuel to the fire, companies today are seeking to add various AI capabilities into this mix; as noted above, Lora Cecere’s admonition on this: “please do not AI this.”
Each of these systems represents best-in-class capabilities for their respective functions. Companies invest in them and achieve significant value in those functions. However, executives continue to lament the lack of breakthrough improvements, or in today’s parlance, the lack of “transformation.” Indeed, a recent PWC survey of executives shows that only 17% of executives believe their supply chain technology investments have yielded expected benefits.
Executives may be expecting transformation, while technology delivers only functional improvements. One might be inclined to think that it is the cumulative bringing together of individual technologies that represents transformation; this is a reasonable statement, but one mitigated by the effort required to assemble these various technologies into a cohesive strategy and architecture, not to mention the ongoing maintenance. Furthermore, integration has historically been used to support linear, sequential business processes. The shortcomings of this are discussed later in this article.
For decades, supply chain processes have been described and diagrammed as linear and sequential. This makes it much easier to describe and understand individual functions. However, physical supply chains and their management processes are non-linear and non-sequential – they are not “chains.”
In 2022, the Association for Supply Chain Management (ASCM) recognized this by releasing a major update to its Supply Chain Operations Reference (SCOR) model (now known as SCOR Digital Standard). The highest-level view of this model is shown in Figure 1. Most notable in this release are two things: 1) the movement to a continuous non-linear, network-like process; and 2) the addition of a level 0 super-process called “Orchestrate.” Orchestration is critical to competitiveness in today’s age in which companies must deliver precision in the face of increasing variability and volatility.
What is orchestration and how do we achieve it in logistics planning and execution? Can we achieve it through integration of existing systems, or do we need something fundamentally different? We explore this question further below and introduce the approach that aThingz has taken to achieve orchestration. But before we do that, let’s discuss how things are done today.
How Do We Do It Today?
Let us now consider an inbound materials scenario for an automotive, industrial, or high-tech electronics manufacturer. This company makes a complex product that is assembled in the US, Canada, and Mexico and sources parts both regionally and from all over the world. Parts, components, and sub-assemblies are transported from source manufacturing plants to assembly plants through a combination of ocean, train, truck, and air transportation assets. The parts and components may also be stored temporarily at various warehouses along the way for buffering, cross-docking, and various other holding purposes.
This inbound logistics environment is supported by and connected to myriad planning and execution software systems. A partial view of the software landscape for inbound materials logistics is shown below in Figure 2.
Forecasted demand and actual orders drive a material requirements planning (MRP) system that explodes the bill of materials for each product; this creates a time-phased plan for each part or component for each assembly plant. These requirements are then fed to a distribution requirements planning (DRP) system (which may be part of the MRP system; both MRP and DRP may be part of an ERP system like SAP). DRP creates a part-by-part time-phased plan of how each part should get from its source manufacturing plant to its destination assembly plant. This considers all lead times, stocking points, and modes of transportation to be used along the way.
Distribution plans are then handed off to transportation planning software (typically part of a transportation management system, or TMS). Transportation planning software considers all constraints, size, packaging, inventory locations, and modes and comes up with where and how parts should be moved from point A to point B to point C and so on. These plans are then provided to a logistics service provider (LSP) or to carriers who then execute the moves (this could be a bidding process or as part of a contract).
So far, this process is one-way and sequential. Enter the “ready to expedite” model of operation. Plans are shoved into the execution realm, where they are executed. But what happens when things go awry in execution? What happens when unusual variability and volatility are encountered? Typically, changes in execution are made and this is then vetted out through end-of-the-month reporting and analytics. Adjustments are then made in the hopes that these issues will not be encountered again. Leading analyst Steve Banker has written a number of articles about this lack of synchronization between planning and execution.
Now overlay on top of these processes and systems the fact that you may have outsourced part or all the planning and management to a logistics service provider (LSP). In this case, the LSP will provide you with detailed reports on how it performed against the plans it has created. But how do you know you are getting the performance to which you are entitled? In many cases, the shipper creates a shadow reporting process to better understand LSP performance, negating some of the cost savings of engaging an LSP.
Performance management is also a critical part of the process. After each cycle (typically a month), customer service (e.g., OTIF) and freight spend are reported. Freight spend is typically done through an offline spreadsheet process in conjunction with the finance department. Attempts are made to reconcile reported spend with invoices and payments. Attempts are also made to understand cost-to-serve through various allocations. However, it is nearly impossible to calculate cost-to-serve at a plant, product, part family or part level. (Cost-to-serve in this context is the landed transportation cost to deliver a part from point of origin to point of consumption).
In the past ten years, visibility platforms have entered the arena of this discussion. These platforms add value by providing near real-time updates on the location and estimated time of arrival of shipments. These platforms have become invaluable in providing shipment status and in identifying problems and attempting to solve them. However, visibility platforms are another system that must integrate to upstream transportation planning systems and downstream warehouse dock systems. So, while they provide much needed visibility, actions needed to resolve problems are left to other systems or to control tower capability that is slapped on top of all these other systems.
And yet after all this work, the logistics planning and execution environment is still rife with various manual techniques, including offline spreadsheet calculations, manual manipulations, phone calls, text messages, emails, and database queries; and of course, the attending latencies.
The bottom line: these systems and the processes they support do not operate as a synchronized orchestra in the sense that the latest SCOR model intends. If agility is the ability to quickly react to variability and volatility and mitigate impact to plans, then this concoction of systems is decidedly unagile.
Can We Integrate Our Way to the Answer?
Can we create an orchestra by integrating the pieces together? Integration allows data to be passed back and forth between the different components and is a large line item in many transformation programs. Historically, integration has represented 30 to 50 percent of the budget of transformation programs and has been a frequent cause of cost overruns and delays.
However, integration does not achieve synchronization[1]; synchronization means that all data that is passed around is foundationally based on the same timestamp. In other words, with integration, there is always systemic latency, and this latency can have hidden compounding effects.
To make this point: Imagine if a neuron in your brain waited an hour to signal another neuron in response to an input. The two neurons are integrated, but they are certainly not synchronized. In other words, synchronization has a time component to it; integration does not. In the context of supply chains, integration is a necessary but insufficient capability. With synchronization, an event occurs, and all constituents are simultaneously notified, at computer clock speed.
In that sense, there is not the idea of components, or separate systems that must be integrated. Systems and processes that were previously separate converge into one. There is one code base simultaneously performing the functions that were previously performed by multiple systems and attachments. (Note: The concept of convergence in the context of logistics was introduced more than a decade ago by Dwight Klappich of Gartner).
Is There a Different Approach?
Fifteen years ago, a similar situation existed in the sales and operations planning (S&OP) process within companies. S&OP is a process executed by the operating committee of a company to balance demand and supply, understand how to handle shortages and excesses, and take advantage of opportunities in the market by adroitly positioning inventory and resources. Historically, this process consisted of disjointed functions like demand planning, supply planning, distribution planning, inventory planning, allocation planning, and order promising. These functions were integrated together through passing data back and forth. Furthermore, when plans were created, they were handed off to execution with reconciliation between actual results and plans happening after-the-fact.
Today, this has all changed. For example, Kinaxis, which is a leader in the S&OP space, can perform all these functions simultaneously through a technique they call “concurrency,” enabled by an in-memory multi-versioning graph database with embedded analytics.
Furthermore, planning and execution are no longer separate, disparate processes. As a plan is executed, threats to the plan are immediately visible and humans and software jointly work together to mitigate or eliminate the threats so that the company can achieve its plans for revenue, costs, and profit. What was once a monthly process is now a continuous process. This is all part of the evolution of supply chain management towards intelligent closed-loop control processes[2].
S&OP for Logistics?
Historically, logistics would go through a monthly process like that of S&OP. A transportation management system (TMS) would create a transportation plan that would then be tendered out and executed. As discussed earlier, there was not a closed, synchronized loop between execution and planning. In recent years, various integrations have happened between visibility platforms, execution systems, and planning; however, as discussed earlier integration is necessary, but not sufficient to deliver the synchronization necessary for today’s need for faster, more intelligent decision making.
At the end of the month, a reconciliation process would be performed, execution would be audited, and payments would be made. As in the case of S&OP, this monthly process is no longer competitive.
Logistics planning and execution must be folded into a single (unified) process that is executed on a continuous forward and backwards (learning) basis. Logistics information (shipment status, locations, ETAs, inventory positions) comes in on a continuous basis, significant variances are alerted to managers, and decisions are made to make adjustments (e.g., change mode, change source, expedite). The plan is updated automatically as part of the adjustment process. Furthermore, the plan includes financial information, so an up-to-the-minute view of financial performance is provided. The transportation plan is a live, accurate representation of what is going to be achieved, as opposed to a point-in-time representation. In this sense, the logistics planning process has evolved from a monthly batch process to a continuous streaming process.
Let us now discuss aThingz, a supply chain software company that has developed a software product, fused with expert services, that delivers such capabilities, driving hundreds of millions of dollars of value.
Enter aThingz Into the Void
Elon Musk is well-known for advocating a “first principles” approach to problem solving. He describes it as a “a physics way of looking at the world.” This approach expressly eschews how things are done today and boils problems down to their key elements and key truths. This approach sometimes leads to completely different solutions to old problems.
aThingz has deconstructed the logistics planning and execution software landscape down to its core elements and reassembled those core elements into a single technology-enabled solution. Instead of multiple systems integrated together in a sequential, linear fashion, aThingz simultaneously synchronizes and orchestrates planning, execution, finance, and performance management into a single continuous process.
This new approach was accelerated and honed as a result of the widespread volatility caused by the pandemic. The need for daily, or even hourly answers to logistics questions could not be met by the fragmented set of systems described in earlier sections of this article. “Airfreight is the only way to keep the plant going. Should I use it, and how much will it cost me?” This type of question needed to be answered on the spot.
In fact, by combining these various elements into a single system and process, aThingz creates a continuous S&OP process for the logistics and transportation problem set. We start with objectives – financials, customer service, asset utilization, we develop a constraint-based plan, we then execute the plan and monitor and manage it as it is being executed.
aThingz calls this process Sales, Logistics Operations Planning with Execution (SLOPE); it is the logistics analog of S&OP. In SLOPE, the plan is a living, breathing entity that slides forward and continually adjusts through time.
Key characteristics of the aThingz solution are outlined in the table below, followed by more detailed descriptions.
FR
Planning and Execution are Indistinguishable
If one were to design a solution following the existing paradigm, there would be a planning system and an execution system. In aThingz, there is just one system. Planning and execution are separate in words only – it is easier to understand how things work when we discuss them as separate entities.
This starts with master data – an always-on data management system that manages, monitors, and synchronizes data elements across different authority domains – shipper (and its various departments – finance, logistics, manufacturing), carriers, logistics service providers and suppliers. This is essentially a control tower for master data.
A logistics operating plan is created and optimized using a combination of heuristics, optimization techniques, and machine learning. The plan considers service-level requirements, costs, inventory, and relevant constraints including modes, capacity, lead times, product size, packaging, and cubing.
The new optimal logistics operating plan is executed by logistics service providers and carriers. Execution information is continuously fed back to the plan through a built-in visibility platform. Deviations from the plan are alerted to planners and managers. Scenarios can then be run to determine the best course of action for mitigating the impact of disturbances.
Demand Forecasting is Built into Logistics Planning
In an inbound materials situation, much of the demand is driven by the MRP system, as described in previous sections. This is the demand for the mainline products (automobile, heavy machinery, electronics, medical equipment). However, there are additional demand streams for SKUs, representing attachments, replacement parts, and other aftermarket products. aThingz provides such forecasting capabilities using a variety of advanced analytics, including machine learning.
Just as important, there is a need to take these independent demand streams and create a dependent multiple-month logistics forecast for such things as trucks, containers, and lanes. This longer-term visibility is valuable to service providers and carriers in effectively managing their business. Furthermore, when orders from the different demand streams change, you can immediately see the change in the logistics forecast.
Operational, Financial, and ESG Metrics are Merged
Operational metrics inform us about things like customer service and cost, as reflected in assets, machinery, labor, transportation, and inventory. Financial metrics give us an understanding of logistics cost-to-serve at multiple levels of granularity, including total, plant, part, and product, and within different time frames from days to weeks to months. ESG metrics are concerned with key environmental, social, and governance issues. Paramount among these is CO2 emissions; since transportation is a significant part of the emissions profile of shippers, it is important to report on this and consider it in decisions.
Historically, supply chain metrics and financial metrics were captured in different systems. There was a back-and-forth integration through spreadsheets, manual manipulations, along with attending latencies. Reconciliation was typically done monthly.
ESG metrics, particularly CO2, have become increasingly important in the last several years. A lot of systems and processes are being built up around ESG. It is important that ESG doesn’t just become another island that has to be integrated into supply chain management.
aThingz has been designed from the bottom up to incorporate operational, financial, and ESG metrics into the same model and to be able to report on them simultaneously. While this may sound easy, it is often non-trivial to integrate operational and financial metrics into the same data model and report on them at multiple levels of detail.
Company and Logistics Metrics are Aligned
It would seem that it would be easy to align the goals and metrics of the company with those of the logistics department. However, consider the following case. The company sees a good short-term opportunity to increase profit by skewing the mix towards higher margin products. That’s the good news.
However, the parts to deliver such mix are not available. The logistics department must expedite and airfreight a number of them. In doing so, they incur higher than expected costs and at the end of the month they miss their budget numbers. The overall company wins; the logistics department loses.
aThingz provides visibility of this situation at the point of decision. In other words, before pulling the trigger on the expedites and airfreight everyone has visibility to the new costs and the new outlook for the monthly cost plan. Logistics becomes part of the solution for the higher profit, versus a drag on it.
Performance Measurement and Management are Continuous
Historically, there is a calendar cadence to measuring and managing operations within enterprises, including logistics operations. Of course, weekly, monthly, and quarterly reporting will continue.
However, the clock speed of business today requires continuous measuring, monitoring, and controlling. aThingz provides the ability to see up-to-the minute logistics performance with a simultaneous view of how the supply chain is performing operationally, financially, and environmentally.
Supply Chain Policy and Structure Decisions are Part of the Process
Changing supply chain structure (e.g., switching source of supply, adding supply) and policies (lead time, inventory stocking) is typically the purview of supply chain design (also called supply chain engineering). This work is usually on a monthly or quarterly basis, or as part of new product programs.
However, as logistics professionals use aThingz to execute their transportation operations, they see immediately how the supply chain is performing. Persistent shifts in demand and supply or persistent underperformance in customer service and costs may have to be addressed immediately. aThingz identifies these opportunities and allows them to be taken through a customer-specific approval process (sometimes called a business transfer approval board).
Continuous Improvement Projects are Directly Identified
The logistics plan is an optimized picture of how transportation should operate within existing constraints. If you are achieving at a level of 70% achievement to your plan, then you are leaving 30% on the table. For a $100M transportation spend, this is $30M; for a $1B spend, this is $300M.
First, various capabilities of aThingz will tighten adherence to the plan. This could mean tens of millions of dollars of value. But what if you are still not achieving full entitlement? Then, changes in execution, policies, or structure must be made.
But how do you know where the improvement opportunities are? aThingz has the answer here as well. As plans are executed, aThingz’ “always-on” autonomous analytics proactively identify the areas where the plan could not be executed, along with the reasons. These areas and the reasons are then cataloged and ranked for review by management. Additionally, this analytics capability also identifies cost avoidance and savings opportunities for future plan optimization. Management can then decide to launch improvement projects around these various areas, in collaboration with partners and carriers.
Data is Continuously Monitored and Managed
Data is the lifeblood of modern supply chain decision systems. Supply chain and logistics processes are multi-organizational, multi-enterprise, and global in nature. Each party contributes data to the decision-making process. Ensuring that all parties are operating against common definitions at all times has been a chronic challenge in the past.
This “data supply chain” is a “hidden supply chain” in many companies. It must be monitored and managed just as the physical and financial supply chains are. aThingz has developed an always-on control tower capability for this data supply chain. It continuously monitors data from all sources and identifies “disturbances,” when data does not align. There is then a remediation workflow to resolve the issue just as there is with a control tower for the physical supply chain.
The “Triple-Double”
aThingz summarizes the above capabilities and differentiators into something they call the “triple-double,” which is shown in Figure 3.
The first part of this is the “triple:” the convergence of the physical supply chain with the informational supply chain, and the financial supply chain. This requires a digital twin that accurately represents the physical supply chain and its policies and can then simultaneously process operational and financial data to make decisions.
The second part of this is the “double:” process convergence across planning and execution. Plans are not point-in-time time-phased representations of what is supposed to happen; they are time phased views of what is supposed to happen, along with what is actually happening.
Together the triple convergence and the double convergence are very powerful.
Visibility + Agility + Adaptability in a Single Logistics Software Solution
Today’s supply chains must deliver increasing precision in the face of increasing variability and volatility. Precision comes in the form of myriad SKU counts, products, and product variants, an array of omni-channel delivery techniques (in-store; buy online pickup in store; curbside service; delivery to location of your choice), and increasingly tight delivery windows. And these are not just B2C requirements; downstream requirements of consumers quickly find their way into tightening B2B requirements.
There are three key capabilities necessary for surviving or perhaps even thriving at the intersection of these forces. These capabilities are highlighted in Figure 4 below.
aThingz unifies these three capabilities.
How is This Possible?
This paper discusses concepts and directions and contrasts them with traditional approaches; it is not intended to be a detailed discussion of technology. However, the reader might ask the question, “what underlying technologies are used to enable these capabilities?” This is a topic for another paper, but a short answer is given here.
aThingz is a cloud-native solution that is built on a composable microservices architecture. It uses the latest technologies for data, data pipeline, analytics, heuristics, optimization techniques, machine learning, and now generative AI. For planning, aThingz has converged heuristics, optimization, and machine learning into a single planning framework and code base. It uses machine learning throughout its various functions including multi-enterprise data management, demand forecasting, monitoring and making recommendations on supply chain structure and policies, and for various elements of the development of feasible, optimized plans.
aThingz is furthermore designed to coexist with existing systems so there is no need for “rip and replace.”
Questions to Ponder
To understand further the key capabilities of aThingz, consider the following questions. These are the types of questions that aThingz can answer within a single software solution.
Can you rapidly run and rerun your logistics plan and compare it against actuals both operationally and financially?
Do you know the criticality of each part for each destination plan? Do you know the inventory positions of these parts?
Do you immediately know your cost-to-serve every time you refresh the system? Do you know it at multiple levels of detail? Can you compare the cost incurred to what’s been invoiced?
Can you keep your data model from multiple sources – shipper, 3PL, supplier – continuously up to date, and immediately identify discrepancies and resolve them?
When you update your logistics plan, are its financials simultaneously updated?
Do you know what your potential savings or entitlement are?
Can you perform root cause analysis of variances to plan?
Can you respond immediately and intelligently to disturbances by running scenarios?
Do you have effective performance management to understand how your 3PL and other service providers are performing?
Do you have visibility into disturbances and is this information unified with your plan?
Can you perform on-the-fly scenarios and immediately see the operational, customer service, and financial impacts of supply chain policy and structure changes?
Can you simultaneously manage inventory, transportation, distribution, and cost in the same system?
Can you do this all in a single system, or does it require a lot of data passing back and forth?
Does your logistics environment rely heavily on phone calls, text messages, emails, offline database queries, spreadsheets, and manual data manipulations?
Do you have to interface with multiple systems?
Show Me the Money
aThingz has designed and developed many of the capabilities discussed in this article by working with some of the world’s largest manufacturing and logistics companies.
But the aThingz journey is not just about delivering good software products and services. More importantly, aThingz is about delivering significant value to clients. And this value can range from tens of millions to hundreds of millions of dollars.
First, aThingz is highly focused on the metrics that are important to the logistics department, and the other organizations it serves, including finance, suppliers, and carriers, as shown in Figure 5.
aThingz’ value proposition impacts various elements of a company’s income statement, balance sheet, and cash flow statement, including revenue, cost of goods sold, profit, inventory, asset utilization, CAPEX, accounts payable, and cash-to-cash cycle.
Typical value delivered is shown in Table 2, below
Adoption of aThingz represents a move away from traditional fragmented approaches to logistics planning and execution towards converged and synchronized approaches. These new approaches have found success in other areas of supply chain management such as S&OP. It’s now time to bring such new approaches to logistics planning and execution.
Footnotes [1] The topic of data synchronization is covered in more detail in a previous article titled “Data Synchronization in Supply Chains.” [2] The concept of a control loop and its application to supply chain management is discussed in more detail in the article “What is a Control Tower?”