Returns and reverse logistics: the blind spot where a micro-automation makes the difference

In recent years, returns have become a more impactful phenomenon than one might think: in fashion e-commerce, return rates often range between 20% and 30%, according to online apparel benchmarks, with increasingly significant operational and economic impacts. According to several industry analyses, handling a return can account for up to 60% of a product’s value, especially when processes are not standardized or supported by automation. Yet, while outbound operations (picking, packing, shipping) have been progressively optimized and automated, reverse logistics continues to be managed in a fragmented, manual, and scarcely measurable way, as if it were an operational exception. As also highlighted by some technology providers, the difficulty goes beyond volumes and lies in the very nature of the process: heterogeneous, unpredictable, and highly dependent on operational decisions.

In this article, you will see where value loss is actually generated in returns management and how to intervene with targeted micro-automations, without resorting to radical warehouse transformations.

Takeaways

• Returns are no longer an exception, but a structural flow with high variability, difficult to manage with traditional models.

• Value is lost throughout the process, between classification, decision-making, and restocking time.

• Traditional automation does not work for returns: targeted micro-automations are needed at decision points and bottlenecks.

• Technologies such as Pick-to-Light, compact sorters (Mushiny, Libiao), and AMRs become effective only if integrated and orchestrated by the WMS.

• Key KPIs (time, errors, and recovered value) improve only with a process-first approach, not with isolated interventions.
  
  
  

Reverse logistics: the most critical (and least controlled) flow

Reverse logistics includes all activities related to goods returning: receiving, identification, verification, decision-making, and restocking or disposal. Often described as the “reverse flow” of the supply chain, it is actually something more complex. Unlike outbound operations, which are predictable, standardized, and optimized for efficiency, returns management is characterized by intrinsic variability:

• volumes are not linear;

• items come back in different conditions;

• return reasons are heterogeneous;

• each item requires a decision.

Even though companies are progressively shifting their focus from returns as a cost to a strategic management logic aimed at value recovery (see Il Giornale della Logistica for more), in operational practice this change is still incomplete. Also in Beta 80’s Smart Logistics models, the return flow is considered a key area where to act with modular, data-driven approaches.

Where value is lost in returns management: end-to-end process analysis

The reverse logistics issue lies in how the return is handled throughout the process, because value is not lost at a single point but along a chain made up of operational steps, decisions, and waiting times. For this reason, to understand where it is really necessary to intervene, the flow must be read in its entirety, from the arrival of the parcel to restocking.

When talking about returns, another important aspect is rarely considered: waste. Reverse logistics is directly linked to sustainability: an inefficient process does not only generate operating costs—products that could be put back on sale end up devalued or out of cycle, with both an economic and environmental impact. Any product that comes back and is not managed correctly risks losing value until it becomes unsellable. Reducing time, making decisions more consistent, and improving traceability means increasing the likelihood that an item becomes available again quickly. In this sense, micro-automations can help prevent errors, reduce waiting times, and keep the process under control, especially at the most critical moments.

Returns classification: the first operational bottleneck

In the most common cases, especially in e-commerce or retail, goods do not arrive in an orderly way: mixed parcels arrive, containing products that differ by type, size, and SKU. This immediately creates an operational issue: before even deciding what to do with the product, parcels need to be made homogeneous. In many warehouses, this phase is still managed manually and requires a sequence of repetitive activities: the operator must identify the item, trace it back to the original order, check the return reason, and decide which flow to route it to (restocking, quality control, scrap). In retail returns, it is often necessary to carry out a real “reverse sorting”: starting from a mixed parcel and rebuilding homogeneous parcels by item code.

At this stage, the most effective micro-automations are not necessarily the most complex ones, but those that can guide the operator and reduce process variability. In practice, this means:

• immediately recognizing the item through integration between scanning and the WMS;

• automatically routing it to the correct flow (restocking, quality control, processing);

• physically supporting the sorting with Pick-to-Light and Put-to-Light systems, which indicate where to place each product while parcels are being reassembled.

In high-volume contexts, especially in fashion, RFID speeds up the identification of inbound items and simplifies restocking, reducing handling times. The critical point remains the ability to make decisions fast and reliable in the very first phases of the process.

Returns sorting: when volume becomes the problem

When volumes grow, the limit of manual handling becomes evident, and this is where one of the most important micro-automations (even if often underestimated) becomes very useful: automatic returns sorting.

In high-volume contexts such as e-commerce, fashion, and 3PL, solutions like Mushiny 3D Sorter, Libiao T-Sort, and Libiao 3D Sort make it possible to automatically sort products by SKU or category, creating homogeneous parcels ready for the next steps. Compared to traditional sorters, these solutions have key features:

• modularity;

• progressive scalability;

• adaptability to existing (brownfield) environments.

The sorter helps speed up and make the returns volume manageable; however, it is important to position it correctly because it is not a universally valid solution, but becomes relevant when:

• volumes are high;

• SKU variety is high;

• manual classification no longer scales.

Alongside Mushiny and Libiao, other players are also working on similar solutions, such as GreyOrange and Geek+, confirming the trend that is consolidating returns sorting as a key point of lightweight automation.

Quality check: the limit of automation

After sorting, you reach the most critical point: quality control. This is the phase in which you decide whether the product will be restocked, refurbished, or scrapped, and it is also the point where, today, automation has the least impact. In e-commerce and fashion contexts, the quality check remains predominantly manual: it is the operator who has to verify:

• product condition and the presence of damage;

• compliance;

• authenticity.

This is precisely because the assessment is often qualitative and contextual, and difficult to standardize with current technologies. That is why, for this phase, it is more effective to rely on control tools such as digital workflows that guide the inspection, structured checklists by product category, and decision tracking in the WMS.

Value here comes from decision consistency: if you are effective in reducing variability, the recovered value increases directly.

From sorting to stock: time is value

Once classified and checked, the product must become available for sale again, and time is the discriminating variable. Any delay between processing and restocking ties up capital, reduces the likelihood of selling, and increases the risk of markdowns or unsold stock. Often, this phase is underestimated and managed manually or with infrastructure that forces operators into rigid workstations (roller conveyors). This is where more flexible technologies such as AMRs (Autonomous Mobile Robots) come into play, dynamically connecting the processing area to storage. After sorting, AMRs:

• transport bins or parcels to stock;

• eliminate idle time;

• reduce dependence on fixed routes.

Unlike AGVs, which follow predefined routes and are poorly suited to variable flows, AMRs move autonomously and dynamically, adapting in real time to operational needs. This flexibility is what makes them particularly effective in returns management, enabling a drastic reduction in throughput time.

Micro-automation in reverse logistics: a process-first approach

At this point, it clearly emerges that reverse logistics is a process design issue more than a technology issue. The technologies exist (Pick-to-Light, compact sorters, AMRs), but their impact depends on where they are placed within the flow. To understand how to do it, the process-first approach is useful: it essentially means starting from three questions and then intervening exactly at those points, without trying to automate the entire warehouse:

• where is time being lost?

• where is information being lost?

• where is a critical decision being made?

In the case of returns, this translates into optimizing the different flows with a logic of:

• • reducing errors and variability in the classification phase;

  • making volume scalable in sorting;

  • making decisions consistent during the quality check;

  • reducing throughput time in the restocking phase.

WMS and orchestration: the real enabler of return management

One element is common to all the micro-automations seen: none of them works effectively without a system that coordinates them, the WMS.

In the context of reverse logistics, the WMS, in addition to tracking stock, becomes the operational brain that:

• recognizes the return and links it to the order;

• decides the flow (restock, inspection, scrap);

• assigns destinations (sorting, stock, processing);

• coordinates the technologies (Pick-to-Light, sorter, AMR).

Software is what determines the behavior of machines, and it is also what allows these solutions to have a “dual value”; for example, the same sorter can handle shipments or returns depending on operational priorities, working on different flows (outbound and reverse) at different times of the day, maximizing the investment.

KPIs in reverse logistics: measuring time, errors, and value

One of the most common limitations in returns management is precisely the lack of granular operational KPIs; often only the overall cost is measured, but not how and where that cost is generated.

The most relevant KPIs in reverse logistics are directly linked to the phases described:

• return processing time: measures overall efficiency;

• classification time: impacts inbound backlog;

• restocking time: impacts recovered value;

• percentage of items returned per order: indicates upstream issues;

• quality check outcomes: measure the quality of decisions.

However, experience shows that some indicators, such as cost per return or operator productivity, are often not managed by the WMS, but by Labor Management systems.

Reverse logistics is one of the most critical areas of the modern supply chain, but also one with the greatest room for improvement; the blind spot lies mainly at the design level rather than the technology level, because a proper approach that “saves” and recovers the value of the return is often missing. Traditional automation, designed for linear flows, struggles to respond to this complexity; micro-automation, instead, works because it is:

• targeted: it intervenes at critical points;

• flexible: it adapts to process variability;

• integrated: it is orchestrated by software.

And it is precisely in this balance between process, technology, and orchestration that reverse logistics becomes a lever for value recovery rather than a cost center.