Robots in Logistics: What Parcel Centres Show About Automation

Robots pay off in logistics where volumes are high and processes are standardised, and the organisation around the technology decides whether they succeed. Sorting in parcel centres has been automated for decades, and the difficult steps sit at the edges of the plant: unloading, singulation and loading. Executives can use this to judge which steps in their own warehouse or dispatch operation suit robots and which conditions must be settled first.

What do parcel centres show about automation?

Parcel centres show that automation spreads step by step and stays away longest where objects are irregular. Volume is the starting point: according to the 2024 courier, express and parcel study of the German federal parcel and express logistics association, providers in Germany carried around 4.2 billion shipments in 2023, and employed roughly 260,000 people to do so. At such volumes every second saved on a manual handling step counts, which is why sorting was mechanised early. Conveyors and cross-belt sorters read the code on each parcel and send it to the right destination within seconds.

The steps that stayed manual for so long are the interesting ones. A container or swap body full of loose parcels has to be unloaded, each parcel has to be singulated and placed on the belt with its code readable, and roll cages and vehicles have to be loaded at the end. Parcels differ in size and weight, from cartons to polybags, and after transport they lie in changing positions. Robot systems that combine gripping and vision have only become products in this area in recent years. In October 2023 the International Federation of Robotics reported that transport and logistics robots were the largest group of professional service robots in 2022, with about 86,000 units sold, an increase of more than 40 per cent on the year before. The market is growing where skilled staff are scarce and peaks occur.

The history of warehouse robots adds a second lesson. When a large online retailer acquired Kiva Systems in 2012 for 775 million US dollars, the point was a principle: shelves or totes travel to the person, and nobody walks the aisles any more. Goods-to-person remains the most commercially successful use of mobile robots in warehouses, because it attacks walking time, which makes up a large share of working time in order picking. The article on order picking explains the effects in intralogistics in detail.

Which logistics and dispatch steps can robots automate?

Robots can automate first those steps that have high piece counts, low variety and a clearly defined start and end point. The decision tool below sets out the criteria. It rates each process step on its own, because a warehouse consists of many steps with very different suitability.

The table explains why sorting was the first step automated in parcel centres: the parcel has a code, the route is fixed and the volumes are high. Unloading fails on the position of objects, singulation on their variety. Newer systems are aimed at precisely these two steps, which is why robots in logistics and dispatch remain an investment topic. Mid-sized warehouses can apply the same logic on a smaller scale: automate the routes first, then the handovers, and the gripping of irregular goods last.

When does a robot pay off in logistics?

A robot pays off in logistics when the annual relief exceeds the cost of capital and operation and the plant runs enough hours to deliver it. The calculation has five lines and can be filled with the figures of your own site. The following example is illustrative and shows the logic with assumed values.

1. Annual relief: saved working hours times fully loaded hourly rate. Example: 6,000 hours times 35 euros gives 210,000 euros.
2. Annual additional cost: maintenance, software, electricity, training and spare parts. Example: 40,000 euros.
3. Net relief: 210,000 minus 40,000 gives 170,000 euros a year.
4. Investment: plant, integration with warehouse management and conveyors, conversion work, commissioning. Example: 600,000 euros.
5. Payback period: 600,000 divided by 170,000 gives about 3.5 years.

Two assumptions decide the outcome. The first is utilisation: the 6,000 hours only materialise if the step actually occurs in the planned volume and the staff either use the freed time elsewhere or temporary labour and overtime fall away. The second is integration, which in projects regularly costs more than the machine. Manufacturers give reliable quotations only after a site survey, so the calculation belongs at the start of the project and is confirmed with real values after the survey. The article on the payback of robot cells shows how such a calculation is built for a single cell.

What must organisation and leadership provide?

Organisation and leadership must decide before the plant arrives who operates it, who is accountable when it fails and how shift work and skills will change. Parcel centres work because technology and maintenance are managed together with shift supervision and planning as one system. A sorting plant has someone responsible for availability and maintenance windows, and a defined manual mode for faults. Anyone who places a robot in an existing process without clarifying these roles meets the same pattern as in every project: the prototype impresses, and operations fail on responsibilities.

Then there is the role of the people at the workplace. Staff whose work changes take over supervision, fault clearing and quality assurance. This requires training and a description of the new tasks before the plant goes live. The article on human-robot collaboration looks at working with robots day to day. Works agreements and occupational safety are matters for specialists. This article provides an overview and does not constitute legal advice.

One decision is often underestimated: the plant needs a measure that shows success in operation beyond the acceptance test. Throughput per hour and availability as a percentage of planned operating time are common, together with the share of parcels that need manual rework. If throughput falls below plan in the first quarter, the rework rate usually shows whether object variety or data quality is the cause.

How does a parcel centre differ from a mid-sized warehouse?

A parcel centre differs from a mid-sized warehouse in volume, standardisation and organisation, and these differences determine what transfers.

For mid-sized companies this leads to a clear approach: start with the step that best meets the criteria in the table and plan the fallback at the same time. Driverless systems for transport are often the entry point because routes and handovers are defined. The economics of these systems are covered in the article on automated guided vehicles.

Introducing robots in logistics with Niehues Consulting

Niehues Consulting is a consultancy for AI management and organisation design from a delivery perspective, based in Düsseldorf. Founder Marin Niehues brings new technology into operations, in projects from VC-funded scale-ups to DAX 40 groups and in more than 20 countries on six continents. Niehues Consulting is not a robotics integrator and does not supply plants. The work comes before and alongside: the business case and the introduction, together with delivery management, that is, the question of who brings what into operation by when.

The approach follows the Niehues Delivery Model with its three phases of Understand, Build and Introduce. Three of its test questions matter for automation projects in logistics: Which business result is to change measurably? Is the business case confirmed with real values? Who is responsible for operation from day one? The entry point is a free initial consultation of 30 minutes, which you can arrange via the contact page.