Why Your Supply Chain Needs More Than ‘Visibility’ – It Needs Orchestration

So, what is logistics orchestration anyway? 

Visibility. Risk Management. Operating system. Decision intelligence. Exceptions management. Product Release. Traceability. Control tower. What’s the difference? 

We hear these words (and their respective functionalities) being co-mingled around in the industry, which makes it awfully confusing for prospective companies and buyers looking to adopt and implement technologies into their network. 

So I’ve put together a guide as to how we see the supply chain tech landscape. 

But first, put your filters on. I want to be clear that we’re solely zoomed in around the intersection of two specific focus areas:

            1.) the Transportation function, and

            2.) the shipment of perishable (cold chain) and high-value product. 

That’s it -- that's our world. Pharma. Produce. Electronics. Oil & Gas. Specifically focused on the movement from point A to Z. No order management. No inventory or warehouse management. No capacity load matching. Just transportation… because we believe transportation – done correctly – is a hard enough problem towards which to channel our entire focus. 

All of this being said, I’ll first define how we view each category of providers, the strengths and weaknesses of each category, and then explain how it all fits together via Logistics Orchestration. 

Devices 

Devices are tracking & sensor modules made up of both active IoT (real-time) and passive data loggers (not real-time). A passive device will solely capture a temperature reading and time stamp with each reading, while an active device will offer some combination of location (via GPS, WIFI, triangulation, etc.), temperature, humidity, shock & vibration, orientation and light. 

Devices are typically attached at the product or pallet level to offer better and more granular visibility (in near real-time) as compared to traditional carrier milestones, which are terribly unreliable. 

Strengths: 

• Devices provide an enablement layer to collect granular visibility data and unlock more advanced capabilities (prediction, recommendation) 

• Competition – because hardware is getting more commoditized and there is more competition both domestically and from LCCs (low cost countries), there is an immense amount of investment to drive form factor, cost and data reliability forward, and be the first to deploy a ‘sticker’ or ‘label’ that is actually functional in Pharma / produce 

• Intervention – devices enable the ability to intervene and salvage product before product loss occurs. Pharma loses $40B worth of product annually, while produce is north of $750B – not to mention the downstream costs of labor (root cause analysis, corrective actions, customer churn, replacement product, etc.) – devices are the root of enabling the reversal of these losses 

Weaknesses: 

• Not data agnostic – each provider is only capable of showing their own device data. Visibility platforms are not setup to bring in other transportation data, such as carrier milestones or ELD data – and overlay it atop of the device data to 1.) fill in gaps, 2.) supplement the dataset, and 3.) identify discrepancies towards source of truth 

• Separation of powers – because they’re not agnostic, there’s no separation of powers – a third-party body that is objectively able to communicate device performance and opportunities for improvement 

• Intelligence & analytics – because hardware is so capitally-intensive, intelligence – such as the ability to predict OTIF failures, or prescribe recommended actions that are contextual in nature and operationally fit within existing workflows – is limited 

When you need IoT devices:  

 Traditionally, you’ll want to leverage IoT within transportation networks that require significant quality control, security or compliance.  

This typically comes in the form of temperature-sensitive products which are subject to quality damage, or time-sensitive or valuable product which may be the target of cargo theft.  

Devices provide a more granular visibility layer that moves from container-level tracking and down to the pallet / product level, making it especially valuable for multi-mode, multi-leg LTL shipments.  

SOP Management & Lane Qualification 

Most common in Pharma and the Life Sciences, lane qualification is a process by which new lanes are qualified through IQ (installation qualification), OQ (operational qualification) and PQ (performance qualification) steps to ensure that the combination of variables (packaging, thermal life, carrier, IoT device, routing, etc.) are adequate to withstand real-world shipping and storage conditions throughout the supply chain. It also includes a requalification process anytime that an existing standard operating procedure or distribution risk assessment is updated / modified. 

There are few tools / functionalities that exist in this space, and PAXAFE just introduced its own. 

 Strengths: 

• Centralized communication, version control and process streamlining 

• Some semblance of risk-scoring 

• Integrations and partnerships with LSPs to facilitate data share 

Weaknesses: 

• Manual SOP / DRA entry – no mass ingestion capabilities 

• Risk scoring is often not representative of real-world performance, and does not account for a Pharmaceutical company’s own pre-approved risk scoring methodologies 

• No dynamic scenario analysis of risk scoring (what happens is I swap out packaging A for packaging B, etc.) without redoing a full risk assessment process 

• No continuous monitoring after lane is qualified – how do I know that changes to seasonality, SOPs, carrier performance, etc. 90, 180, 360 days out has not impacted this qualification? 

When you need Lane Qualification:  

If you are in the Pharmaceutical business, lane qualification and risk assessments are a vital compliance checkmark to assure regulatory bodies that your company took the requisite steps to de-risk its supply chain. 

 
Companies will engage in lane qualification activities such as performance mapping, thermal modeling and distribution risk assessments typically when 1.) they are setting up a new product on a new lane, or 2.) they need to revalidate a product / SOP.  

 Other industries have less formal and sophisticated qualification processes, but they’re typically more so about establishing a ‘performance baseline’ of what is ‘normal’ rather than focusing on compliance.  

Real Time Transportation Visibility (RTTV)

This contingent of companies focus on delivering what I call ‘visibility plus’ to their customers. Their primary focus is building out a breadth of integrations to consolidate a broad spectrum of data all into one view / platform, and offer additional value-add capabilities (e.g. capacity load matching, warehouse management, security monitoring, etc.) 

Strengths:  

• This segment of companies houses some of the more mature companies that have been around a bit longer, and the companies that are objectively valued at higher valuations 

• Breadth of capability: These aggregators will often times consolidate other capabilities that shippers want, such as capacity load planning, warehouse & yard management, etc.  

• Good monitoring & exceptions management capabilities for live shipments 

• Strong intelligence capabilities around ocean visibility & ocean time of arrival 

 Weaknesses:  

• Data quality is poor – typical coverage includes 70-80%, making data not actionable, which undermines the value & ROI 

• Limited capabilities around temperature management & workflows, particularly at the product / pallet level

• Limited aggregate intelligence capabilities towards planning & optimization, limiting larger buckets of savings

When you need RTTV:  

Shipments of commodities, low value freight (e.g. clothing, retail) where container-level tracking and ocean & rail tracking are at the forefront of importance. 

Industries with high detention / demurrage penalty fees.  

Warehouses with large yards, where inbound product could get lost / damaged. 

Industries where freight is booked in the spot market.  

Performance & Risk Management: 

Companies pay $20, $30 and even $40+ per data logger or IoT device / shipment! That’s a lot of money to let that data go to waste. 

Plus – here’s a dirty little secret: the majority of your value from real-time visibility programs won’t come in the form of product loss reduction or intervening on a live shipment. 

It will come in the form of planning and optimizing parameters to reduce cost and waste to begin with. 

It will come from having the data and confidence to switch from Air to Ocean on a particular lane. Or to reduce Contracted Lead Time with an LSP, saving on inventory carrying costs. 

It will come from reducing over and under-packaging on each lane. 

It will come from selecting the right device / LSP / packaging company to begin with, and continuously monitoring their performance over time to drive accountability, trust and mutual investment 

It will come from bridging the gap between static SOPs and risk assessments, and having a live feedback loop to continuously monitor and update these processes as your performance and risk within your network evolves. 

There are also few providers in this space. 

Strengths: 

• Strong focus on supplier mapping and ‘black swan’ events and external risk – weather, geopolitical, etc. 

• Attempts to integrate transportation data with other supply chain data (e.g. inventory management, order management, etc.) but not yet operational 

• General aggregate analytics and dashboards exist to support querying of specific questions 

Weaknesses: 

• Limited focus on every day decisions that drive the majority of OTIF failures (packaging, shipment time, routing, carrier selection, SOP management, etc.) 

• Lack of auto-geofencing (the ability to map out routes and waypoints automatically) from a lane (origin <> destination pairing) prevents hyper-contextualization and pairing of external risk data with IoT / transportation data 

• Intelligence is touted, but much of it is done behind the scenes manually – which sparks buyer doubt when results are either slow, expensive or inaccurate 

• No focus on internal SOPs, which prevents context and understanding of existing workflows 

When you need Performance & Risk Management:  

Have you ever heard the phrase ‘insanity is doing the same thing over and over again, but expecting a different result?’ 

That’s kind of how most industries treat the supply chain today.  All of the focus is around exceptions management and reducing the issue NOW…TODAY…RIGHT THIS MOMENT. All about monitoring and exceptions management! 

Without much foresight to tomorrow…  

 The ability to aggregate large amounts of data – from IoT & carrier milestones, to SOPs, to weather and geopolitical – and then understand the patterns and correlations around that data will enable your planning and quality teams to make better decisions tomorrow, and – over time – reduce the amount of issues that your monitoring teams will need to respond to tomorrow.  

 Companies need aggregate performance and risk management when they reach a critical mass of data or decisions. This inflection point varies by industry.  

 For Pharma, this might mean 50 unique lanes (origin < > destination pairs), or 1,500+ shipments per month.  

For Produce, this might mean 5000 monthly loads.  

Generative AI 

Large Language Models (LLMs) and AI agents have been transformational in how we collect, analyze, process and query data, and take action on that data. 

Most shippers and BCOs are coming to the realization that the value that comes out of these LLMs is not in the infrastructure layer, but rather in the application layer – where these models are fine tuned to specific areas of expertise (e.g. cold chain transportation).  

The make vs. buy decision is still top of mind, where companies that go the BUY route are really buying into the RAG infrastructure, agent setup and model fine-tuning; while companies that try to leverage LLMs directly have to figure out how to minimize hallucinations and contextualize data outputs of a generic large language model. 

  LLM Strengths: 

• Transformational tools that have the potential to fundamentally change how data is consumed and actioned within the organization 

• Tremendous automation potential 

Weaknesses: 

• LLM models change quickly – it’s expensive to stay up to date in understanding which models are optimal towards which kinds of queries 

• Expensive data teams required to build & maintain RAG infrastructure, train & retrain models 
• Generic LLMs are not well equipped for tailored, fine-tuned applications 
  

This brings us to the magic phrase of the day: Logistics Orchestration.  

 Logistics Orchestration is the ability to bring multiples of the aforementioned capabilities together into one platform – to be able to support both the strategic planning and operational aspects of a transportation network.  

 This is not to be confused with broader supply chain orchestration, where leaders like Kinaxis and E2Open focus on stringing together much broader functional datasets (order management, inventory planning and management, sales & operations, etc.) but not go as deep into any specific module (e.g. transportation). 

Strengths:  

• Typically data agnostic, with the ability to consolidate capabilities to provide buyers with one point of use 

• Third-party trust layer that provides intelligence in a non-bias way 

• Intelligence and analytics capabilities are typically more advanced  

 Weaknesses: 

• The marketing can be stronger than the product capability – overpromising and under-delivering – particularly on the prediction / intelligence / automation front – leaves customers disappointed, particularly when work has to be done manually by large teams behind the scenes vs. the promised software experience 

• Reliant on external data sources with limited control over quality – ‘garbage in, garbage out’  

• Lots of ‘frenemies’ – providers must work together to deliver an optimal seamless experience for the customers, but are often times reluctant to do so due to perception of competition. Most of the time this perception is exacerbated, but there are instances where future roadmaps do converge and collaboration becomes painful 

The PAXAFE difference