Maze Environment Research Articles

An Improved Dyna-Q Algorithm Inspired by the Forward Prediction Mechanism in the Rat Brain for Mobile Robot Path Planning.

The traditional Model-Based Reinforcement Learning (MBRL) algorithm has high computational cost, poor convergence, and poor performance in robot spatial cognition and navigation tasks, and it cannot fully explain the ability of animals to quickly adapt to environmental changes and learn a variety of complex tasks. Studies have shown that vicarious trial and error (VTE) and the hippocampus forward prediction mechanism in rats and other mammals can be used as key components of action selection in MBRL to support "goal-oriented" behavior. Therefore, we propose an improved Dyna-Q algorithm inspired by the forward prediction mechanism of the hippocampus to solve the above problems and tackle the exploration-exploitation dilemma of Reinforcement Learning (RL). This algorithm alternately presents the potential path in the future for mobile robots and dynamically adjusts the sweep length according to the decision certainty, so as to determine action selection. We test the performance of the algorithm in a two-dimensional maze environment with static and dynamic obstacles, respectively. Compared with classic RL algorithms like State-Action-Reward-State-Action (SARSA) and Dyna-Q, the algorithm can speed up spatial cognition and improve the global search ability of path planning. In addition, our method reflects key features of how the brain organizes MBRL to effectively solve difficult tasks such as navigation, and it provides a new idea for spatial cognitive tasks from a biological perspective.

ARCHIGENIUS - VR MAZE SOLVING USING A* ALGORIHTM

Game is one of the entertainment media that is rapidly growing. On this day, playing games is one way to spare time for refreshing. One of the popular games this year is an action game such as First-Person Shooter (FPS). Besides, the FPS game genre has developed rapidly to make it more interesting for the player. Non-player Character (NPC) has an essential function in artificially and automatically to make a game more immersive and consisting of various kinds. Also, from observation, adding challenges when playing, namely enemy characters, can make a player more enjoyable. Besides, one of modern technology to develop innovations for supporting a game is called virtual reality. So, in this research, the FPS game is built by adding NPC elements with virtual reality technology. Also, the gameplay in this game is a survival maze. Based on the evaluation, this game application can display objects in a virtual reality smoothly, and all feature in this game can display and run effectively. Besides, the respondents also give good feedback for user testing which has average score 4.66 from several users. So, the conclusion from this research is this game can make the FPS game more interesting for users by adding NPC, 3D element, and virtual reality. Keywords - First-person Game, Virtual Reality, Survival Maze, Non-Player Character, virtual maze environment.

Towards Personally Relevant Navigation: The Differential Effects of Cognitive Style and Map Orientation on Spatial Knowledge Development

Under emergencies such as floods and fires or during indoor navigation where cues from local landmarks and a Global Positioning System (GPS) are no longer available, the acquisition of comprehensive environmental representation becomes particularly important. Several studies demonstrated that individual differences in cognitive style might play an important role in creating a complete environmental representation and spatial navigation. However, this relationship between cognitive style and spatial navigation is not well researched. This study hypothesized that a specific type of map orientation (north-up vs. forward-up) might be more efficient for individuals with different cognitive styles. Forty participants were recruited to perform spatial tasks in a virtual maze environment to understand how cognitive style may relate to spatial navigation abilities, particularly the acquisition of survey and route knowledge. To measure survey knowledge, pointing direction tests and sketch map tests were employed, whereas, for route knowledge, the landmark sequencing test and route retracing test were employed. The results showed that both field-dependent and field-independent participants showed more accurate canonical organization in their sketch map task with a north-up map than with a forward-up map, with field-independent participants outperforming field-dependent participants in canonical organization scores. The map orientation did not influence the performance of Field-Independent participants on the pointing direct test, with field-dependent participants showing higher angular error with north-up maps. Regarding route knowledge, field-independent participants had more accurate responses in the landmark sequencing tests with a north-up map than with a forward-up map. On the other hand, field-dependent participants had higher accuracy in landmark sequencing tests in the forward-up map condition than in the north-up map condition. In the route retracing test, however, the map orientation had no statistically significant effect on different cognitive style groups. The results indicate that cognitive style may affect the relationship between map orientation and spatial knowledge acquisition.

Research on autonomous mobile robot maze navigation problem based on Dijkstras algorithm

In recent years, the field of autonomous mobile robotics has garnered significant attention due to its potential applications in various domains such as logistics, surveillance, and search and rescue operations. A crucial challenge in this area is the efficient navigation of robots within complex and dynamic environments, particularly when navigating through maze-like structures. The maze navigation problem involves finding optimal paths for robots to traverse from their initial positions to designated destinations while avoiding obstacles and making intelligent decisions to ensure timely and safe navigation. This study aims to investigate and apply Dijkstras algorithm to solve the maze navigation problem for autonomous mobile robots. By analyzing the navigation challenges faced by autonomous mobile robots in maze environments, a solution based on Dijkstras algorithm is proposed. In conclusion, this study contributes to the field of autonomous mobile robotics by proposing and evaluating the application of Dijkstras algorithm for maze navigation. The experimental results validate its potential to address the challenges of navigating intricate maze environments. However, it is acknowledged that further refinement and innovation are possible to continue improving the performance of autonomous mobile robots in maze navigation scenarios.

Ultrasonic sensor decision-making algorithm for mobile robot motion in maze environment

An autonomous mobile robot is one that can move from one location to another without the intervention of a human. A maze environment is a complex environment since it contains many obstacles and the major problem is moving through it. To avoid obstacles while moving through the maze, the mobile robot must be designed with an algorithm. This work proposes a decision-making system with an ultrasonic sensor to allow the developed autonomous mobile robot to avoid obstacles in any maze setting through its movements. The maze was designed with a size of 100×200 cm2 for the case study. Due to the height dimension of the barrier (20 cm) and the height dimension of the mobile robot including the ultrasonic sensor (20 cm), a distance of 20 cm was taken between the wall (obstacle) and the sensor. The result distance between the (wood wall) object and the sensor indicates that it is a reasonable distance chosen for the mobile robot to move and turn with flexibility as it travels through the maze environment from 0 cm to 300 cm. This mobile robot path took 1 minute to finish at a speed of 5 cm/sec, indicating that it is a quick algorithm as compared with related work.

An optimized Q-Learning algorithm for mobile robot local path planning

The Q-Learning algorithm is a reinforcement learning technique widely used in various fields such as path planning, intelligent transportation, penetration testing, among others. It primarily involves the interaction between an agent and its environment, enabling the agent to learn an optimal strategy that maximizes cumulative rewards. Most non-agent-based path planning algorithms face challenges in exploring completely unknown environments effectively, lacking efficient perception in unfamiliar settings. Additionally, many Q-Learning-based path planning algorithms suffer from slow convergence and susceptibility to getting stuck in local optimal solutions. To address these issues, an optimized version of the Q-Learning algorithm (Optimized Q-Learning, O-QL) is proposed and applied to local path planning of mobile robots. O-QL introduces novel Q-table initialization methods, incorporates a new action-selection policy, and a new reward function, and adapts the Root Mean Square Propagation (RMSprop) method in the learning rate adjustment. This adjustment dynamically tunes the learning rate based on gradient changes to accelerate learning and enhance path planning efficiency. Simulation experiments are carried out in three maze environments with different complexity levels, and the performance of the algorithm in local path planning is evaluated using steps, exploration reward, learning rate change and running time. The experimental results demonstrate that O-QL exhibits improvements across all four metrics compared to existing algorithms.

FF-RRT*: a sampling-improved path planning algorithm for mobile robots against concave cavity obstacle

The slow convergence rate and large cost of the initial solution limit the performance of rapidly exploring random tree star (RRT*). To address this issue, this paper proposes a modified RRT* algorithm (defined as FF-RRT*) that creates an optimal initial solution with a fast convergence rate. An improved hybrid sampling method is proposed to speed up the convergence rate by decreasing the iterations and overcoming the application limitation of the original hybrid sampling method towards concave cavity obstacle. The improved hybrid sampling method combines the goal bias sampling strategy and random sampling strategy, which requires a few searching time, resulting in a faster convergence rate than the existing method. Then, a parent node is created for the sampling node to optimize the path. Finally, the performance of FF-RRT* is validated in four simulation environments and compared with the other algorithms. The FF-RRT* shortens 32% of the convergence time in complex maze environment and 25% of the convergence time in simple maze environment compared to F-RRT*. And in a complex maze with a concave cavity obstacle, the average convergence time of Fast-RRT* in this environment is 134% more than the complex maze environment compared to 12% with F-RRT* and 34% with FF-RRT*. The simulation results show that FF-RRT* possesses superior performance compared to the other algorithms, and also fits with a much more complex environment.

0468 Effects of apnea-induced SWS disruption with and without supplemental oxygen on overnight spatial memory in CPAP-treated OSA

Abstract Introduction Sleep disturbance from obstructive sleep apnea (OSA) impairs overnight memory processing. This may be mediated by changes to slow wave sleep (SWS) as less slow wave activity (SWA) is associated with worse overnight spatial and declarative memory. We developed a within-subjects model of SWS-specific CPAP-withdrawal to create 3 polysomnologically (PSG)-verified conditions of 1) stable-SWS on CPAP, 2) SWS-fragmentation with intermittent hypoxemia (OSAsws), and 3) SWS-fragmentation with reduced hypoxemia (OSAsws+O2). Using these conditions, we sought to dissociate of effects of SWS disruption and intermittent hypoxemia on overnight spatial navigational memory. Methods We included 33 patients (7 female, average age 54yrs) with moderate-to-severe OSA (AHI4%>20/hr) who were CPAP-treatment adherent. During each PSG visit, they explored one of three 3D maze environments, performing 3 timed trials before and after each of 3 randomized and counterbalanced SWS disruption conditions [1)CPAP, 2)OSASWS & 3)OSASWS+O2]. %SWS, OSA severity during SWS, frontal SWA(0.5-4 Hz), and overnight change in maze completion times were compared according to PSG condition using Friedman Rank Sum and Conover’s tests. The relationship between SWA and maze performance was tested using Pearson correlation coefficient. Results %SWS was significantly greater during CPAP compared to CPAP-withdrawal conditions (CPAP=17.2%, OSAsws=13.3%, OSAsws+O2=13.8%, p< 0.02). Apnea hypopnea index (AHI3A) in SWS during CPAP-withdrawal was significantly greater compared to CPAP (CPAP=0/hr, OSAsws=24.4/hr, OSAsws+O2=19.3/hr, p< 0.00001). Hypoxic burden in SWS was significantly reduced in the OSAsws+O2 condition vs OSAsws (23 %min/hr vs 37 %min/hr, p < 0.001). No significant differences were observed between the three conditions for SWA or overnight change in maze completion times. When including all PSG visits, greater SWA was significantly associated with better overnight maze improvements (R=0.28,p< 0.01) which approached significance for the CPAP condition in isolation (R=0.33,p=0.068). Conclusion SWS-specific CPAP-withdrawal resulted in recurrence of mild OSA in SWS; supplemental oxygen reduced the hypoxic burden, but SWA did not differ between OSA with hypoxemia and without. There was, however, an overall positive relationship between SWA and maze performance. These observations suggest that the effect on spatial navigation of hypoxia in OSA is mediated by disruption of SWA. Support (if any) R01AG056682, AASM Focused Project Award.

The Occipital Place Area Is Recruited for Echo-Acoustically Guided Navigation in Blind Human Echolocators.

In the investigation of the brain areas involved in human spatial navigation, the traditional focus has been on visually guided navigation in sighted people. Consequently, it is unclear whether the involved areas also support navigational abilities in other modalities. We explored this possibility by testing whether the occipital place area (OPA), a region associated with visual boundary-based navigation in sighted people, has a similar role in echo-acoustically guided navigation in blind human echolocators. We used fMRI to measure brain activity in 6 blind echolocation experts (EEs; five males, one female), 12 blind controls (BCs; six males, six females), and 14 sighted controls (SCs; eight males, six females) as they listened to prerecorded echolocation sounds that conveyed either a route taken through one of three maze environments, a scrambled (i.e., spatiotemporally incoherent) control sound, or a no-echo control sound. We found significantly greater activity in the OPA of EEs, but not the control groups, when they listened to the coherent route sounds relative to the scrambled sounds. This provides evidence that the OPA of the human navigation brain network is not strictly tied to the visual modality but can be recruited for nonvisual navigation. We also found that EEs, but not BCs or SCs, recruited early visual cortex for processing of echo acoustic information. This is consistent with the recent notion that the human brain is organized flexibly by task rather than by specific modalities.SIGNIFICANCE STATEMENT There has been much research on the brain areas involved in visually guided navigation, but we do not know whether the same or different brain regions are involved when blind people use a sense other than vision to navigate. In this study, we show that one part of the brain (occipital place area) known to play a specific role in visually guided navigation is also active in blind human echolocators when they use reflected sound to navigate their environment. This finding opens up new ways of understanding how people navigate, and informs our ability to provide rehabilitative support to people with vision loss.

Nanorobots target search strategies using swarm intelligence

Controlling a swarm of bee-like nanorobots in search of a target in a human body-like environment is a challenge. The challenge lies in the causal factors of emergence. This research focused on designing and developing a bee-inspired algorithm to provide instructions to coordinate simple and naïve nanorobots in search of a target. Experiments were conducted through simulations to determine the rate at which myriads of agents converge at a target. A maze environment depicting human blood vessels was used to deploy the nanorobots. Different sample population sizes were used, and the results indicate that the bigger the sample size, the higher the convergence speed. Compared with a single complicated robot, the swarm of intelligently controlled nanorobots proved efficient and effective in finding a target in a human-like environment.

Flexible Behavioral Decision Making of Mobile Robot in Dynamic Environment

How to achieve flexible behavioral decision making in a dynamic environment is an important prerequisite for a mobile robot to execute various tasks. Traditional methods suffer from discontinuous and unstable learning ability. This article brings the mobile robot continuous and stable learning ability in the dynamic environment, through improving the binding manner between the cerebellum and basal ganglia in the neuromodulatory system proposed in our previous work, thus to enhance its flexible behavioral decision-making performance. Moreover, a more biological significance index, i.e., the curiosity index, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$Cur$ </tex-math></inline-formula> , is designed to mimic the activity switch of the locus coeruleus between the tonic and phasic mode, to modulate the exploration-exploitation tradeoff in the reinforcement learning (RL) of the basal ganglia. Influence of varying learning rate and varying discount rate on the performance of the RL is also investigated. The experiment results in static and dynamic environments, as well as the comparative experiment in the complex maze environment and the complex dynamic environment, demonstrate the potential of the proposed neuromodulatory system.

Spatio-temporal categorization for first-person-view videos using a convolutional variational autoencoder and Gaussian processes.

In this study, HcVGH, a method that learns spatio-temporal categories by segmenting first-person-view (FPV) videos captured by mobile robots, is proposed. Humans perceive continuous high-dimensional information by dividing and categorizing it into significant segments. This unsupervised segmentation capability is considered important for mobile robots to learn spatial knowledge. The proposed HcVGH combines a convolutional variational autoencoder (cVAE) with HVGH, a past method, which follows the hierarchical Dirichlet process-variational autoencoder-Gaussian process-hidden semi-Markov model comprising deep generative and statistical models. In the experiment, FPV videos of an agent were used in a simulated maze environment. FPV videos contain spatial information, and spatial knowledge can be learned by segmenting them. Using the FPV-video dataset, the segmentation performance of the proposed model was compared with previous models: HVGH and hierarchical recurrent state space model. The average segmentation F-measure achieved by HcVGH was 0.77; therefore, HcVGH outperformed the baseline methods. Furthermore, the experimental results showed that the parameters that represent the movability of the maze environment can be learned.

A Brain-Inspired Model of Hippocampal Spatial Cognition Based on a Memory-Replay Mechanism.

Since the hippocampus plays an important role in memory and spatial cognition, the study of spatial computation models inspired by the hippocampus has attracted much attention. This study relies mainly on reward signals for learning environments and planning paths. As reward signals in a complex or large-scale environment attenuate sharply, the spatial cognition and path planning performance of such models will decrease clearly as a result. Aiming to solve this problem, we present a brain-inspired mechanism, a Memory-Replay Mechanism, that is inspired by the reactivation function of place cells in the hippocampus. We classify the path memory according to the reward information and find the overlapping place cells in different categories of path memory to segment and reconstruct the memory to form a “virtual path”, replaying the memory by associating the reward information. We conducted a series of navigation experiments in a simple environment called a Morris water maze (MWM) and in a complex environment, and we compared our model with a reinforcement learning model and other brain-inspired models. The experimental results show that under the same conditions, our model has a higher rate of environmental exploration and more stable signal transmission, and the average reward obtained under stable conditions was 14.12% higher than RL with random-experience replay. Our model also shows good performance in complex maze environments where signals are easily attenuated. Moreover, the performance of our model at bifurcations is consistent with neurophysiological studies.

Autonomous UAV Path-Planning Optimization Using Metaheuristic Approach for Predisaster Assessment

In this article, different state-of-the-art metaheuristic algorithms are analyzed to incorporate the collision-free path-planning approach for UAVs in a disaster situation. The efficient path planning is used to identify the main cause of disaster like bushfires that are badly affecting the forest ecosystem throughout the globe. This novel approach is a first step toward a predisaster assessment and possibilities to save the survivors in minimal time. Different metaheuristic algorithms, such as PSO, GWO, WOA, BMO, and DGBCO, are compared for UAV path optimization capability. In order to test the robustness of our proposed model, four different scenarios are presented which include general environment, condense environment, maze environment, and dynamic environment. In each scenario, the obstacles are placed in such a way to increase the overall path complexity for a UAV to reach the destination. The exploration and exploitation groups working simultaneously in a dynamic environment make it effective for UAV path planning using the DGBCO algorithm. Based on the parameters selected, the DGBCO algorithm outperforms other algorithms and achieves 24.5% less transportation cost and 13.3% less computational time. Hence, DGBCO can be efficiently applied for UAV path-planning optimization in any of the aforementioned environments.

Path discovering in maze area using mobile robot

Robotic maze pathfinding problems deal with detecting the correct route from the start point to the end-point in a virtual maze environment consisting of walls. Automated robot mobility is a significant feature, which enables a mobile robot to traverse a maze independently, from one position to another, without human intervention. There is a myriad of autonomous industrial mobile robot applications, including the transportation of goods and parts, domestic cleaning, indoor security surveillance, airport baggage couriering, and a plethora of other applications to traverse dangerous locations. This paper proposes a pathfinding mobile robot in a virtual maze based on a combination of a simplified left-hand algorithm and a line-following control algorithm. The mobile robot works in any maze to determine a route from the initial starting point to the end-point. The approach outlined in this paper uses a left-hand algorithm to solve the maze problem and a line-follower control algorithm to enable the robot to move in a straight line through the virtual maze. The algorithm used is less complicated and prevents the robot from falling into infinity loops compared to the traditional wall-follower algorithm.

Forgetting Enhances Episodic Control With Structured Memories.

Forgetting is a normal process in healthy brains, and evidence suggests that the mammalian brain forgets more than is required based on limitations of mnemonic capacity. Episodic memories, in particular, are liable to be forgotten over time. Researchers have hypothesized that it may be beneficial for decision making to forget episodic memories over time. Reinforcement learning offers a normative framework in which to test such hypotheses. Here, we show that a reinforcement learning agent that uses an episodic memory cache to find rewards in maze environments can forget a large percentage of older memories without any performance impairments, if they utilize mnemonic representations that contain structural information about space. Moreover, we show that some forgetting can actually provide a benefit in performance compared to agents with unbounded memories. Our analyses of the agents show that forgetting reduces the influence of outdated information and states which are not frequently visited on the policies produced by the episodic control system. These results support the hypothesis that some degree of forgetting can be beneficial for decision making, which can help to explain why the brain forgets more than is required by capacity limitations.

An Improved Dueling Deep Double-Q Network Based on Prioritized Experience Replay for Path Planning of Unmanned Surface Vehicles

Unmanned Surface Vehicle (USV) has a broad application prospect and autonomous path planning as its crucial technology has developed into a hot research direction in the field of USV research. This paper proposes an Improved Dueling Deep Double-Q Network Based on Prioritized Experience Replay (IPD3QN) to address the slow and unstable convergence of traditional Deep Q Network (DQN) algorithms in autonomous path planning of USV. Firstly, we use the deep double Q-Network to decouple the selection and calculation of the target Q value action to eliminate overestimation. The prioritized experience replay method is adopted to extract experience samples from the experience replay unit, increase the utilization rate of actual samples, and accelerate the training speed of the neural network. Then, the neural network is optimized by introducing a dueling network structure. Finally, the soft update method is used to improve the stability of the algorithm, and the dynamic ϵ-greedy method is used to find the optimal strategy. The experiments are first conducted in the Open AI Gym test platform to pre-validate the algorithm for two classical control problems: the Cart pole and Mountain Car problems. The impact of algorithm hyperparameters on the model performance is analyzed in detail. The algorithm is then validated in the Maze environment. The comparative analysis of simulation experiments shows that IPD3QN has a significant improvement in learning performance regarding convergence speed and convergence stability compared with DQN, D3QN, PD2QN, PDQN, PD3QN. Also, USV can plan the optimal path according to the actual navigation environment with the IPD3QN algorithm.

Efficient and High Path Quality Autonomous Exploration and Trajectory Planning of UAV in an Unknown Environment

The ability of an autonomous Unmanned Aerial Vehicle (UAV) in an unknown environment is a prerequisite for its execution of complex tasks and is the main research direction in related fields. The autonomous navigation of UAVs in unknown environments requires solving the problem of autonomous exploration of the surrounding environment and path planning, which determines whether the drones can complete mission-based flights safely and efficiently. Existing UAV autonomous flight systems hardly perform well in terms of efficient exploration and flight trajectory quality. This paper establishes an integrated solution for autonomous exploration and path planning. In terms of autonomous exploration, frontier-based and sampling-based exploration strategies are integrated to achieve fast and effective exploration performance. In the study of path planning in complex environments, an advanced Rapidly Exploring Random Tree (RRT) algorithm combining the adaptive weights and dynamic step size is proposed, which effectively solves the problem of balancing flight time and trajectory quality. Then, this paper uses the Hermite difference polynomial to optimization the trajectory generated by the RRT algorithm. We named proposed UAV autonomous flight system as Frontier and Sampling-based Exploration and Advanced RRT Planner system (FSEPlanner). Simulation performs in both apartment and maze environment, and results show that the proposed FSEPlanner algorithm achieves greatly improved time consumption and path distances, and the smoothed path is more in line with the actual flight needs of a UAV.

Subchronic impacts of 2,4-D herbicide Weedestroy®AM40 on associative learning in juvenile yellow perch (Perca flavescens).

Aquatic herbicides are commonly used to control a wide variety of invasive and nuisance plants. One common active ingredient used in commercial herbicide formulations in Midwestern states is 2,4-dichlorophenoxyacetic acid (2,4-D). Due to the stability of 2,4-D in aquatic environments, many non-target aquatic species experience prolonged exposure throughout critical developmental life stages that can affect essential behaviors. However, the impacts of 2,4-D exposure on learning behaviors in juvenile fish are poorly understood. Therefore, we conducted a series of experiments using a maze environment to determine the effects of a commercial 2,4-D amine salt herbicide formulation (Weedestroy®AM40; WAM40; at 0.00, 0.50, 2.00, and 50.00 mg/L 2,4-D acid equivalent (a.e.)) exposure on juvenile yellow perch's ability to perform a feed associated learning behavior. We observed a significant decrease in the ability of yellow perch to correctly complete the feed associated learning behavior within 200 s when exposed to WAM40 at 2.00 and 50.00 mg/L 2,4-D as compared to controls (p=0.0002; p < 0.0001, respectively) and within 600 s when exposed to WAM40 at 2.00 and 50.0 mg/L 2,4-D as compared to the controls (p=0.0107 and p < 0.0001). These data suggest that exposure to 2,4-D in WAM40 can both increase the amount of time it takes for yellow perch to complete a feed associated learning behavior and/or obstruct the behavior altogether. Further experiments showed no significant decreases in locomotion (p > 0.05), hunger motivation (p > 0.05), and a visually guided startle response (p > 0.05), in all treatment groups tested as compared to controls. This suggests that 2,4-D in WAM40 does not inhibit feed associated learning behaviors via interaction with these mechanisms. Altogether, the results indicate that the use of 2,4-D herbicides for weed control in aquatic ecosystems could present risks to cognitive functions that control essential behaviors of yellow perch.

Measuring Intelligence in Natural and Artificial Systems

A systematic understanding of the relationship between intelligence and consciousness can only be achieved when we can accurately measure intelligence and consciousness. In other work, I have suggested how the measurement of consciousness can be improved by reframing the science of consciousness as a search for mathematical theories that map between physical and conscious states. This paper discusses the measurement of intelligence in natural and artificial systems. While reasonable methods exist for measuring intelligence in humans, these can only be partly generalized to non-human animals and they cannot be applied to artificial systems. Some universal measures of intelligence have been developed, but their dependence on goals and rewards creates serious problems. This paper sets out a new universal algorithm for measuring intelligence that is based on a system’s ability to make accurate predictions. This algorithm can measure intelligence in humans, non-human animals and artificial systems. Preliminary experiments have demonstrated that it can measure the changing intelligence of an agent in a maze environment. This new measure of intelligence could lead to a much better understanding of the relationship between intelligence and consciousness in natural and artificial systems, and it has many practical applications, particularly in AI safety.