Published on: March 12, 2025 / update from: March 12, 2025 - Author: Konrad Wolfenstein
Researcher Sepehr Samavi and Prof. Angela Schoellig next to robots Jack - Image: Astrid Eckert, Muenchen
Pioneering work in robotics: TUM develops forward -looking robots
Autonomous systems: How to learn robots to interact with people
In a world that develops rapidly in the direction of automation and artificial intelligence, autonomous systems are an increasingly important part of our everyday life. From self -driving cars to intelligent assistance robots to highly developed industrial plants - the ability of machines to make self -employed decisions and act in complex environments transforms numerous areas of our life. A particularly exciting and challenging discipline within the robotics is the development of systems that can move safely and efficiently in dynamic environments populated by humans. This is not just about avoiding obstacles, but also about understanding, predicting and reacting the behavior of people in order to ensure smooth and safe interaction.
Exactly at this interface of robotics, artificial intelligence and human behavior, researchers from the renowned Technical University of Munich (TUM) work high pressure. In your Learning Systems and Robotics Lab, under the direction of Professor Angela Schoellig, you have developed an innovative robot called “Jack”, which is able to navigate with remarkable skill and foresight through crowds. What distinguishes Jack from many other robots is his ability to not only perceive the immediate surroundings, but also to actively think about how people will move around and how they could react to their own movements. This foresighting way of thinking enables Jack to plan his way through lively rooms not only reactively, but also proactively and intelligently.
Suitable for:
The challenge of navigation in crowds
Navigation in crowds is an enormous challenge for robots that goes far beyond simple obstacle avoidance. In contrast to static or predictable environments, crowds are dynamic, unpredictable and characterized by complex social interactions. Everyone in a lot moves individually, but at the same time influences the movements of others. This interdependence, combined with the natural variability of human behavior, makes it extremely difficult for robots to move safely and efficiently.
Traditional navigation algorithms for robots, which are often based on rigid rules and simple sensor data, quickly reach their limits in such environments. They usually react react to obstacles by abruptly stopping or dodging, which can lead to undesirable traffic jams, inefficient routes or even dangerous situations in a crowd. In order to successfully move in crowds, robots therefore need a much more progressive form of intelligence, which enables them to understand human behavior, to predict and actively involve their navigation planning.
Jack's innovative approach: forward -looking thinking and interaction
The robot Jack developed by the TUM researchers goes a decisive step beyond traditional approaches. His core is a sophisticated algorithm that enables him not only to perceive the movements of people in his area, but also to actively predict and to involve his own route planning. Professor Schoellig emphasizes the fundamental difference to conventional methods: “Our robot modeled how people will react to his movement to plan his own way. That is the big difference to other approaches that typically ignore this interaction. ”
This ability to model interaction is the key to Jack's success. Instead of only considering people as unpredictable obstacles, Jack sees her as an intelligent actor whose behavior he can sometimes predict and even influence. This enables him to move through crowds that resemble human navigation in many ways. He does not hesitate to move in gaps, anticipates the movements of pedestrians and adapts his route dynamically to avoid collisions and at the same time efficiently achieve his goal.
Sensor and computing power in interaction
In order to cope with this demanding task, Jack is equipped with highly developed sensors and computing power. A central element is a lidar sensor (Light Detection and Ranging), which permanently sends laser beams into the area and receives the reflected signals. From this data, the lidar creates a precise 360-degree card in the environment in real time, which not only captures static objects, but in particular also the position and movement of people. The lidar thus provides the robot a detailed “picture” of its surroundings, which forms the basis for its navigation decisions.
In addition to the lidar, Jack has sensors in his bikes, which precisely measure his own pace and the distance covered. This information is crucial to precisely determine your own position in the area and optimize the efficiency of the navigation. All sensor data are processed by a powerful on -board computer that is able to carry out complex algorithms in real time. This computer is the “brain” of Jack and responsible for the analysis of the sensor data, the prediction of human movements and the calculation of the optimal route.
Suitable for:
The algorithm in detail: prediction, planning and adaptation
The heart of Jack's intelligence is the navigation algorithm developed by the TUM researchers. This algorithm works in several steps to enable Jack to ensure safe and efficient navigation in crowds.
1. Perception and data acquisition
Initially, Jack continuously collects data about his surroundings with the help of his sensors. The lidar provides information about the position and movement of people, while the wheel sensors provide data on the robot's own movement.
2. Prediction of human movements
Based on the data collected, the algorithm analyzes the movement pattern of the people in the area. He tries to predict the likely paths that people will take over in the next seconds. This prediction is based on statistical models that have been learned from extensive data records of human movement behavior in crowds.
3. Route planning
At the same time, the algorithm plans the optimal route to the goal of the robot. He not only takes into account the predicted movements of people, but also the robots' own skills and restrictions, such as its speed and maneuverability. The goal is to find a route that leads to the goal as quickly and efficiently as possible without risking collisions with people.
4. Dynamic adaptation
A central aspect of the algorithm is its ability to adapt dynamically. The entire process of data acquisition, prediction and route planning is continuously repeated about ten times a second. This allows Jack to adapt his route to the constantly changing environment in real time. This high adaptation frequency is essential to navigate in a dynamic environment with many people safely and efficiently, since the robot recognizes the ways of people at the same time and reacts to how the TUM researcher Sepehr explains Samavi.
Learning from human behavior: the key to human -like navigation
Another crucial aspect of Jack's intelligence is his ability to learn from human behavior. The TUM researchers did not simply program Jack with rigid rules and algorithms, but gave him the opportunity to continuously improve through the analysis of data of human movement behavior.
Professor Schoellig explains that the mathematical model on which the planning algorithm is based was derived from human movements and translated into equations. The algorithm is therefore not based on abstract assumptions about human behavior, but directly on real data that document the movements of crowds. In order to enable this, the researchers collected extensive data records that describe human behavior in different situations and environments and serve as teaching material for Jack.
By analyzing this data, Jack learns to recognize, anticipate typical movement patterns of people and to involve his own decisions. For example, he learns that people usually dodge when they are heading for an obstacle or that they adapt their speed to avoid a collision. These findings flow into the algorithm and enable Jack to behave in a way that resembles the intuitive behavior of people in crowds.
A concrete example of this learning process is Jack's handling of potential collisions. A traditional robot would usually stop immediately as soon as he recognizes an obstacle, such as a person, on a collision course. Jack, on the other hand, who has learned from human behavior, reacts more differently. He also calculates that people will usually adapt and dodge to avoid a collision. Therefore, he does not stop immediately, but continues his movement, while at the same time observing the reaction of man. Only if there are signs that people will not dodge do Jack plan at short notice and choose an alternative route. This behavior is much more efficient and more human -like than the abrupt stop of a traditional robot.
Evolutionary development: from reactive too interactive
The development of Jack's navigation skills was an evolutionary process that went into three stages. Each level represents progress in the complexity and intelligence of algorithm.
Level 1: Reactive navigation.
In the first stage, Jack only reactively reacted to his surroundings. He evaded obstacles as soon as he perceived them without predicting or anticipating the behavior of people. This stage was functional, but inefficient and often led to abrupt stops and detours.
Level 2: Predictive navigation.
In the second stage, the algorithm was expanded to predict the movement of oncoming people. This made it possible Jack to navigate more forward -looking and avoid collisions before they were imminent. This level was already a significant progress, but was still limited because it largely ignored the interaction between robots and humans.
Level 3: Interactive navigation.
The current version of Jack represents the third and most advanced level of evolution: interactive navigation. In this level, Jack is not only able to predict the movements of people, but also to actively take into account how people will react to their own movements. He is able to influence people's behavior through his own behavior and at the same time avoid collisions. This interactive ability is the crucial breakthrough that makes Jack a really intelligent and human -like navigation system.
Researcher Samavi explains that Jack can predict other people's movements on the one hand and at the same time is able to influence their actions through his own behavior while avoiding collisions. This form of interactive navigation enables Jack to move safely, efficiently, socially acceptable and intuitively through crowds.
Areas of application: from delivery robots to autonomous driving
The innovative technology that is in Jack has enormous potential for a variety of application areas. Although Jack was initially developed as a research platform, the TUM researchers are already thinking about concrete possible uses in the real world.
Delivery robot
A close application are delivery robots that can autonomously deliver goods and packages in urban environments. These robots must be able to move safely and efficiently on sidewalks, in pedestrian zones and in lively city centers. Jack's ability to navigate in crowds is of crucial importance for this. In the future, autonomous delivery robots could make a significant contribution to solving problems of the “last mile” in logistics and relieving urban traffic.
Suitable for:
Wheelchairs
Another promising application is the integration of technology into intelligent wheelchairs. Navigation in lively environments can be a major challenge for people with mobility restrictions. A wheelchair that is equipped with Jacks navigation algorithm could significantly improve the independence and quality of life of these people. The wheelchair could automatically avoid obstacles, move safely through crowds and bring the user autonomously to the desired destination.
Autonomous driving
Professor Schoellig sees autonomous driving as a particularly relevant field of application for interactive navigation technology. It emphasizes that these interactive scenarios are a central challenge. In complex traffic situations, for example when threading on motorways, when turning to crossings or when dealing with pedestrians and cyclists, it is essential not only to plan your own movement, but also to predict the behavior of other road users and to include them in their own planning. The ability of the technology for interactive navigation could thus make a significant contribution to the development of safe and efficient autonomous vehicles. As an example, it leads to threading on a highway: When a vehicle drives on the acceleration gauge of a motorway entrance, many drivers coming from behind change tracks or braking slightly. It is precisely in such situations that the new approach enables the reactions of the other road users to be adequately taken into account.
Humanoid robots
Humanoid robots could benefit particularly from the algorithms, especially in areas such as care, service or production in which they work closely with people. In order to be used and effective by humans, it is essential that they can navigate safely and intuitively in human environments. However, Professor Schoellig refers to a central challenge: While a moving robot can simply stop if necessary, humanoid robots are currently still quite unstable and quickly lose their balance. The improvement of the stability of humanid robots in dynamic environments represents an important field of research that must be further developed in order to make the full potential of interactive navigation also usable for humanoid robots.
Advanced robot navigation: As Jack understands human behavior
The research of TUM in the area of interactive robot navigation represents significant progress on the way to intelligent and autonomous systems that can act safely and efficiently in human surroundings. The robot Jack impressively shows that it is possible to develop machines that not only perceive their surroundings, but also understand human behavior, predict and include them in their decisions. This ability to interactive navigation opens up new opportunities for a variety of applications, from delivery robots to intelligent wheelchairs to autonomous driving.
The development of Jack is only the beginning. Research in the field of robotics and artificial intelligence is progressing rapidly, and we can expect further exciting innovations in the coming years and decades. The integration of robots into our everyday life will become increasingly natural, and autonomous systems will play an increasingly important role in our society. It is therefore of crucial importance that we make the development of these technologies responsible and take into account the ethical and social aspects from the start. This is the only way we can ensure that robots and people can work together for the benefit of everyone in the future.
Suitable for:
Your global marketing and business development partner
☑️ Our business language is English or German
☑️ NEW: Correspondence in your national language!
Konrad Wolfenstein
I would be happy to serve you and my team as a personal advisor.
You can contact me by filling out the contact form or simply call me on +49 89 89 674 804 (Munich) . My email address is: wolfenstein ∂ xpert.digital
I'm looking forward to our joint project.
