Robotics. For some, that might still sound like science fiction, but in our warehouses and supply chains, it's long been a reality. Robotics is massively transforming how we work in logistics.To shed some light on this, we had a real expert as our guest: Victor Splittgerber from WAKU Robotics. He is deeply rooted in the field of robotics for logistics and has recently become increasingly involved in software solutions for robotics. In our latest podcast episode, he gave us some exciting insights that you don't get everywhere.
Why Robotics is Booming in Logistics Right Now
Victor made it clear: Hardly any area in robotics is developing as fast as logistics. When we talk about robots for our industry, it's primarily about mobile robots. These are the autonomous devices that move on wheels. Amazon has been demonstrating this for years, with nearly a million of them soon to be in use there, according to Victor. But we're seeing more and more of them here too, because the technology is getting better and more affordable at the same time.
Ultimately, the business case decides, of course. Without it, no automation will pay off. But mobile robotics offers enormous potential and disrupts various processes in the warehouse:
- Order Picking: Robots take over walking distances and replace classic systems. This saves enormous time and effort.
- Sorting: Often small robots, reminiscent of vacuum cleaners, handle parcel sorting. They are cost-effective and perfect for facilities that have many destinations but perhaps don't need extremely high throughput.
- Transport: Think of autonomous forklifts or automated guided vehicles (AGVs) that transport pallets from goods reception to the warehouse. Such fixed routes can be wonderfully automated.
In addition, segments like cleaning or inventory in the warehouse are growing strongly. Victor described an example of huge towers that drive through the warehouse every night and record inventory. This way, you always have an overview and minimize errors – something that would be difficult to achieve manually with this density and precision.
AI in Robotics: Between Hype and Hard Reality
Of course, the topic of Artificial Intelligence in connection with robotics is on everyone's lips. Also in our conversation. People talk about AI getting a physical body. This is supposedly going to revolutionize all areas of life in the future.
What are the drivers?
- Hardware Power: Chips are getting faster and faster, and computing power is growing exponentially. What we see today is just a foretaste of what's to come.
- Image Recognition: Robots are getting better at understanding their environment. They can recognize people and objects, interact with them, and even understand speech. Imagine being able to simply tell a robot: "Bring me pallet X to gate Y!" A lot is happening here right now.
- Navigation and Simulation: AI helps robots find their way, plan optimal routes, and simulate processes in virtual environments. The results of these simulations can then be directly transferred to the robots.
But now comes the important point: the reality.
For me, this was the most exciting part of the conversation. Victor gave us a very realistic but essential assessment: Many of the breakthroughs in AI research still need a lot of time before they are truly ready for practical use. He aptly compared it to ChatGPT: Theoretically, you can do anything with it, but the practical results are often not what you imagine.
Many robots currently in use still operate rather statically and "limited" when compared to what would theoretically be possible. We are still a good way from human-level interaction. Even if many startups advertise with "a lot of AI," the actions the robot performs are often still relatively simple. There are still limitations in recognizing or picking items, or only certain process steps can be mapped. A human can (still) do a bit more, is more flexible.
The core message here: Currently, AI is not yet the decisive factor that determines the success or failure of a robotics project. This will change in a few years, then you won't be competitive without AI. But at the moment, many established technologies that are less talked about – such as classic automated guided vehicles (AGVs) – are much more usable, robust, and predictable.
Humanoid Robots: A Look Behind the Hype
The topic of humanoid robots is currently at the peak of public attention. Think of Boston Dynamics, Optimus from Elon Musk and Tesla. A humanoid is simply a robot that looks human – two legs, two arms, and a head. The idea behind it: Our world is made for humans, so a human-like robot can best move within it and take on various tasks. In addition, there is an enormous amount of data (such as videos of human movements) on which such robots could be trained.
Leading companies in this segment have raised gigantic sums of capital in a very short time. However, experienced roboticists view the whole thing with healthy skepticism. They know how protracted the development from prototypes to production readiness is and how difficult it is to generate a viable business case. Although companies like GXO Logistics, Inc. or Mercedes-Benz AG are already experimenting with humanoid robots in small pilot projects, their productivity is currently still far from that of a human.
A clear recommendation from Victor in this situation: As a logistics or production company with little experience in robotics, let alone humanoids, you should NOT intensively deal with humanoid robots NOW. It's still too early, and it brings little added value that one hopes for. There are so many easier automation projects that deliver immediate added value and are based on proven technology – think of AutoStore™, picking systems, or sorting systems. It will be some time until humanoid robots are truly "off the shelf" and meaningfully usable in everyday life.
Human and Robot: How Collaboration in the Warehouse Works
The use of robots changes the role of many logisticians. It's moving more and more from direct employee management to managing automation. This brings new tasks that many don't even consider today:
- Spare Parts Management: An often underestimated topic. How do I efficiently manage spare parts to minimize downtime?
- Knowledge Management: If a robot shows an error, it must quickly be clear what the message means and how to fix it. This is a gap that Victor fills with WAKU Care, a maintenance software. For example, many sensors are sensitive to dust – if dust accumulates, the robot stops. Without the necessary knowledge of how to clean the sensors, the machine stands still. If this knowledge is missing because it is not documented or employees have left the company, you are constantly reinventing the wheel. Not very efficient for an automation case.
Humans will continue to play a crucial role in the future. A robot always needs someone to operate it – the "Robot Operator." These people take care of maintenance, assist robots with malfunctions, or perform cleaning tasks. This means new job profiles are emerging, while others are shifting.
In warehouses that still operate largely manually today, automation will increase significantly, for example, through autonomous forklifts. But there will always be people to supervise these machines. Supervision ratios will be defined, for example, one person for a certain number of robots.
The Future of Robotics: What to Expect (and What Not)
Victor assumes that robotics will become a standard feature, similar to how computer chips are found in almost every device today. Every new forklift will have the ability to drive autonomously. This will establish itself much faster in warehouses than on public roads because the warehouse environment is more controllable and employees can be trained. Fewer variables than, for example, in public road traffic.
The overarching goal: To handle warehouse tasks with significantly fewer people. This is also urgently necessary given the demographic change in many regions.
However, Victor is skeptical about developments in public spaces, such as autonomous vehicles, drones, or delivery robots on sidewalks, at least in the EU. The new EU AI Act makes the use of AI-based systems in public areas extremely complex. In China or the USA, you already see thousands of autonomous vehicles, but in our region, this will probably remain an exception for the foreseeable future.
Victor's Appeal to the Robotics Industry and Users
If Victor had one wish, it would be more realism in the approach.
- Adapt processes: It is essential to rethink and adapt existing warehouse processes if you want to introduce robotics. A robot can do some things better, but others worse than a human. You cannot simply replace a human worker 1:1 with a robot and expect everything to run smoothly.
- Patience and realism: Robotics projects take time. There are many examples of failed projects because they were too fast. It often takes half to a full year for systems to reach their full productivity. In the beginning, they can be supervision-intensive, and productivity may still be lower. It's a learning process for everyone involved; you have to know that and factor it in.
- Use independent consulting: Victor strongly advises companies with little experience in robotics to work with external consulting firms. There are an incredible number
