New Algorithm Helps Robots Escape Dead Ends

AI brings us convenience in the daily life and allows us to eat food delivered by robots without leaving home. Although the food delivery robots seem smart and capable, they also have obvious shortcomings, many of the robots are not good at escaping blind ends and avoiding obstacles.

Researchers from the Institute of Engineering and Technology at the Nicolaus Copernicus University have developed an algorithm that will allow mobile robots to escape blind alleys and avoid obstacles.

The principle of the algorithm comes from the human brain. Why don’t people walk into a dead end in a complex building? Because people have the ability to perceive, that is, when we see icons and signs, we can assess whether there are obstacles in front of us. This flexibility of the human brain allows us to recognize the direction at all times when we walk on the street, instead of walking to the end of the street.

The human brain allows us on the right path without wasting time looking back, so can robots also do it?

“This is exactly the behavior we based on when developing the path planning algorithm that allows mobile robots to avoid dead ends” explains Prof. Tomasz Tarczewski, from the Institute of Engineering and Technology of the NCU.

“Nowadays mobile robots are becoming more and more common in warehouses and production halls. Despite special routes, master algorithms coordinating the work of many robots in one hall, it may happen that the path of the robot will be blocked. The main problem is that they often run into so-called dead ends.” Prof. Tomasz Tarczewski said.

The way to solve this problem is to perform local path planning algorithms to avoid various types of obstacles, the algorithm makes it possible for the robot to automatically unblock itself and continue its route.

“The principle operation of most of these mechanisms involves the introduction of additional algorithms to guide the robot out of a dead end. In this way, the distance traveled by the mobile robot increases due to changing routes, causing unnecessary power consumption,” explains Rafal Szczepanski. “For this, we used lidar sensors, or laser scanners, that provide information on the robot’s distance from obstacles.”

Then, based on these readings, the research team developed a terminal prediction mechanism. The robot compares real-time information about the available space with its own size and makes a passability decision based on the information. In order for a robot to give up exploring an impassable part, it must be equipped with mechanisms to bypass it.

“Generally, it is the creation of additional virtual objects and adding them to reality. This technology is commonly used on smartphones, tablets for various applications such as mobile games, photo and video applications, and education toys,” said Artur Bereit.

“It works by creating a virtual wall or an additional barrier to the left or right of the robot when a dead end is detected, preventing it from passing through, to guide it in the right direction,” explains Rafal Szczepanski. “It is this innovative and interdisciplinary combination of mobile robotics and augmented reality that allows us to publish our research.”

“Our research team implemented an augmented reality-supported potential field method development algorithm in the mobile robot Husarion robot 2.0 PRO and conducted a series of laboratory tests to demonstrate the effectiveness of the proposed solution. The results of the study are very promising — we demonstrate a significant improvement in the performance of potential field algorithms commonly used in mobile robotics.” Rafal Szczepanski said.

About Datatang

Founded in 2011, Datatang is a professional artificial intelligence data service provider and committed to providing high-quality training data and data services for global AI companies. Relying on own data resources, technical advantages and intensive data processing experiences, Datatang provides data services to 1000+ companies and institutions worldwide. Datatang entered Chinese stock market (NEEQ: 831428) in 2014 and became the first listed company in China’s artificial intelligence data service industry.

If you need data services, please feel free to contact us:




Off-the-shelf AI training data, on-demand data collection & annotation services

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

The Pros and Cons of AI

An imaginary conversation with Elon Musk at a pub in Yorkshire.

Reading Response_Anatomy of an AI System

Role of AI / ML in Digital Transformation

BCF Ventures & Scale AI Partner to Offer Corporate Innovation Program for Supply Chain Start-ups

AmazeWIT Bangalore

Transforming Patient Experiences with AI

Is a Neural Network of Glass an AI?

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store


Off-the-shelf AI training data, on-demand data collection & annotation services

More from Medium

Google’s New Patent Helps Drivers Keep Attention While Driving

Two Architectures for Real-Time Vanishing Point Detection

Surmounting Steep Challenges in Training and Validating Automated Driving Functions — dSPACE

Paper Review: “EdgeFed: Optimized Federated Learning Based on Edge Computing”