August 27, 2025
The GIST Smarter navigation: AI helps robots stay on track without a map 
Lisa Lock
scientific editor
Robert Egan
associate editor
Editors' notes
This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:
fact-checked
proofread
Navigating without a map is a difficult task for robots, especially when they can't reliably determine where they are. A new AI-powered solution helps robots overcome this challenge by training them to make movement decisions that also protect their ability to localize. Instead of blindly heading toward a target, the robot evaluates the visual richness of its surroundings and favors routes where it's less likely to get "lost."
By combining deep reinforcement learning (DRL) with real-time feedback on pose estimation, this approach enables robots to avoid both collisions and localization failures. The result: more intelligent navigation, better performance in tricky environments, and a major step toward autonomous systems that truly understand their limitations.
Traditional robot navigation methods either require detailed maps or assume accurate localization is always available—assumptions that break down in indoor or unfamiliar environments. Visual simultaneous localization and mapping (SLAM) systems, often used as a fallback, can easily fail in scenes lacking distinct textures or during abrupt movements, leading to severe navigational errors.
Many previous AI models focused only on finding collision-free paths, ignoring whether the robot could stay properly localized along the way. Even those that consider localization often rely on rigid penalty thresholds, unable to adapt to changing conditions. Due to these challenges, a more flexible and awareness-driven navigation strategy is urgently needed for robots to perform reliably across diverse, real-world scenarios.
A research team from Cardiff University and Hohai University has developed a new deep reinforcement learning (DRL) model that helps robots plan smarter, safer paths in complex indoor environments. Their study, published in IET Cyber-Systems and Robotics in July 2025, introduces a localization-aware navigation policy trained with RGB-D camera input and real-time feedback from ORB-SLAM2.
Rather than relying on pre-set thresholds or fixed maps, the robot learns to adapt dynamically to visual conditions in its environment—boosting its ability to maintain localization and significantly improving navigation success rates in simulated tests.
The core innovation lies in integrating localization quality into every navigation decision. The robot is trained using a compact state representation that reflects the spatial distribution of visual map points, partitioned into 24 angular regions around its body. This design helps the robot identify which directions are visually "safer" to travel through—meaning areas are more likely to provide reliable localization data.
In addition, the researchers introduced a new reward function based on Relative Pose Error (RPE), offering instant feedback on whether a particular movement action improves or worsens the robot's understanding of its position. Unlike previous models that used static thresholds, this system employs a dynamic threshold that adjusts in real time, depending on environmental conditions.
To evaluate the approach, the team conducted extensive training using the iGibson simulation environment and tested their model against four baseline methods. In challenging indoor scenarios, the new model outperformed others by a wide margin, achieving a 49% success rate compared to just 33% for conventional SLAM-based navigation.
It also showed significantly lower localization error and better adaptability in new environments. Notably, the model consistently chose longer, but safer routes—demonstrating the value of prioritizing localization robustness over shortest-path efficiency.
"Our aim wasn't just to teach the robot to move—it was to teach it to think about how well it knows where it is," said Dr. Ze Ji, the study's senior author. "Navigation isn't only about avoiding walls; it's about maintaining confidence in your position every step of the way. By integrating perception and planning, our model enables smarter, safer movement in uncertain spaces. This could pave the way for more autonomous robots that can handle the complexities of the real world without constant human oversight."
The implications of this work extend across indoor robotics—from service robots in hospitals and homes to warehouse automation systems. In environments where GPS doesn't work and visual conditions vary, being able to assess and respond to localization reliability is crucial. This method equips robots with the awareness to adjust their strategies based on how well they can see and understand their surroundings.
Looking forward, the team plans to test their model on real robots and in dynamic scenes with pedestrians. With further development, the approach could become a key building block for trustworthy, mapless navigation in real-world human environments.
More information: Yan Gao et al, Deep Reinforcement Learning for Localisability‐Aware Mapless Navigation, IET Cyber-Systems and Robotics (2025). DOI: 10.1049/csy2.70018
Provided by Zhejiang University Citation: Smarter navigation: AI helps robots stay on track without a map (2025, August 27) retrieved 27 August 2025 from https://techxplore.com/news/2025-08-smarter-ai-robots-stay-track.html This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
Explore further
New AI system could change how autonomous vehicles navigate without GPS shares
Feedback to editors
