WiSER-X: Wireless Signals-based Efficient Decentralized Multi-Robot Exploration without Explicit Information Exchange
Ninad Jadhav, Meghna Behari, Robert J. Wood, Stephanie Gil · 2024
Read WiSER-X: Wireless Signals-based Efficient Decentralized Multi-Robot Exploration without Explicit Information Exchange by Ninad Jadhav, Meghna Behari, Robert J. Wood, Stephanie Gil (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-19876.txt.
Introduction
Ninad Jadhav, Meghna Behari, Robert Wood, and Stephanie Gil explore unknown spaces with robots that share no maps, in work from 2024. Teams must cover new ground fast while link bandwidth stays tight in the field. Signal pings alone can give rough partner positions without full data exchange.
Their method uses only local partner range guesses from onboard radio signals. Each robot picks its next goal to cut overlap with nearby partners by that clue. Teams can end at split times with no shared map held in common. Tests cut overlap by over half versus a no sharing baseline method. Hardware runs with full onboard sensing backed the simulation gains.
Catalog record
| CREATOR | Ninad Jadhav, Meghna Behari, Robert J. Wood, Stephanie Gil |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-19876
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
WiSER-X: Wireless Signals-based Efficient Decentralized Multi-Robot Exploration without Explicit Information Exchange Ninad Jadhav, Meghna Behari, Robert J. Wood, Stephanie Gil Abstract We introduce a Wireless Signal based Efficient multi-Robot eXploration (WiSER-X) algorithm applicable to a decentralized team of robots exploring an unknown environment with communication bandwidth constraints. WiSER-X relies only on local inter-robot relative position estimates, that can be obtained by exchanging signal pings from onboard sensors such as WiFi, Ultra-Wide Band, amongst others, to inform the exploration decisions of individual robots to minimize redundant coverage overlaps. Furthermore, WiSER-X also enables asynchronous termination without requiring a shared map between the robots. It also adapts to heterogeneous robot behaviors and even complete failures in unknown environment while ensuring complete coverage. Simulations show that WiSER-X leads to 58% lower overlap than a zero-information-sharing baseline algorithm-1 and only 23% more overlap than a full-information-sharing algorithm baseline algorithm-2. Fig. 1: From left to right, the first panel shows the hardware components used for onboard sensing. The second panel visualizes the environment with each robot’s sensing radius within it. The third panel shows how a robot can leverage both inter-robot positioning and historical exploration to infer which parts of the environment have been explored. Finally, the last panel shows how this information is used to update the utilities of the frontiers for a given robot, and it uses this to guide its exploration. I Introduction Coordinated decentralized multi-robot exploration of an unknown environment is relevant for many applications such as search-and-rescue as it enables faster exploration and avoids single points of failure [ 1 ] . Efficient exploration of the environment traditionally requires information exchange between the coordinating robots. Often it includes local map estimates, common features or landmarks, and relative position estimates. However, such information exchange is often limited by communication bandwidth constraints, non-line-of-sight conditions, remote operations, or on-board computation limitations on robots with size, weight and power constraints. For instance, robots in a search-and-rescue mission may lack line-of-sight, or underwater gliders may operate without GPS and high-bandwidth communication. Existing works have made significant progress to address these limitations by developing methods such as (1) using prior centralized information to pre-assign robots to regions of the environment before deployment [ 2 ] (2) sharing post-processed data like sparse environmental features to estimate common coverage areas [ 3 ] , or (3) requiring periodic rendezvous [ 4 ] , such as surfacing to communicate or obtain GPS fixes during underwater exploration [ 5 ] . While these approaches have enabled significant towards coordinated exploration in challenging environments, they come with limitations such as precluding real-time adaptation to heterogeneous robot performance and failures, and/or extending mission times by requiring rendezvous. In an ideal setting, robots could assess each other’s coverage in real-time without explicit communication, reducing redundancy in exploration while eliminating interruptions or high bandwidth information exchanges. Additionally, the algorithm would adapt to heterogeneous behaviors, for example, assigning larger portions of the environment to be explored by robots with better navigation capabilities. Along these lines, we introduce WiSER-X, a decentralized coordination algorithm that allows robots to improve their local frontier exploration strategy by leveraging onboard relative Angle-of-Arrival (AOA) and range measurements to other robots in the team (Fig 1 ). AOA and range can be measured in a variety of settings, including acoustically in underwater environments [ 6 ] , and using RF signals (such as WiFi and Ultra Wide-band) in indoor environments for example. Importantly, communication pings needed to measure AOA and range between robots are much more lightweight, typically requiring 32 kb/s as opposed to 2.5 Mb/s for full feature map exchange, and can traverse longer distances through NLOS occlusions. We address the following key challenges when developing WiSER-X. As choosing a frontier to visit depends on its utility [ 7 ] , we update the information gain of robots’ frontiers by estimating potential overlaps, computed based on the estimated relative position of neighboring robots obtained from range and bearing sensor measurements. As such, once a robots’ local frontier’s utility falls below a certain threshold, it is marked as invalid and no longer a candidate during the current timestep. However, robots still need to know when to terminate exploration without a shared map. To address this, robots’ maintain a history of the relative positions locally in a hgrid data structure, enabling them to track which robots have visited various areas within the environments boundary (known apriori). Thus a robot terminates exploration asynchronously when there are no valid frontiers left or when it independently estimates that the environment has been sufficiently covered. WiSER-X also adapts to to heterogeneity in robot behavior or even complete failures in real-time, minimizing the exploration time without loss of coverage in comparison to the baseline algorithms. To accurately estimate positions from noisy AOA and range measurements from real sensors, we use an Extended Kalman Filter (EKF) to maintain accurate relative positioning, even in environments with potential multipath interference. WiSER-X also adjusts the weight of a neighboring robot’s information overlap at a frontier based on the certainty of their relative position estimates. WiSER-X thus enables efficient, coordinated decentralized multi-robot exploration using only AOA and range measurements acquired from ping packets at 32 kb/s. We validate our method through extensive simulations and hardware experiments, showing (1) improvements in time to completion, (2) reduced coverage overlap, (3) minimized missed areas in exploration, and (4) effective handling heterogeneity in robot behaviors. II Related Work Several strategies have been proposed to improve exploration of an unknown environment with a single robot or a team of robots [ 8 , 9 , 10 , 11 , 12 , 13 , 14 ] . However, decentralized multi-robot exploration in unknown environments predominantly requires significant data exchange such as map updates, sensor data, trajectories or positions in common frame of reference [ 15 ] . Reducing communication exchange is a significant area of study and much of the research in this area focuses on minimizing such explicit data exchange [ 16 , 17 , 18 ] using methods such as compressing map data before transmission to reduce communication load, reducing the number of robots involved in communication [ 19 ] , limiting data exchange to when robots are in close proximity [ 20 , 21 ] , and designating a subset of robots as connectivity robots that are responsible for maintaining network links [ 22 ] . Intermittent rendezvous, where robots plan periodic meetings to share information throughout a mission, is another common technique [ 23 , 24 , 25 ] . However, all of these methods rely on some baseline form of information transmission which are prone to multiple issues such as perceptual aliasing and not always possible in communication bandwidth constrained environments. Thus, despite these developments in reducing communication load via explicit information exchange for multi-robot tasks, there is still limited research on fully decentralized exploration methods that leverage low-bandwidth implicit information exchange to increase efficiency in exploration. While methods do discuss updating individual robots’ belief states to guide their exploration strategies [ 26 , 27 , 28 ] , they often still depend on non-trivial levels of communication or a centralized “auction” system [ 29 ] . To circumvent such constraints, some methods aim to