A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2), 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the efficiency of heuristic and incremental searches, yet still finds shortest paths. It achieves a speedup of one to two orders of magnitudes( over repeated A [10] searches by modifying previous search results locally. D has been extensively used on real robots, including outdoor HMMWVs [15], including UGV Demo II vehicles as part of the DARPA Unmanned Ground Vehicle program. It is currently also being integrated into Mars Rover prototypes and tactical mobile robot prototypes for urban reconnaissance [5, 9, 17] D is also used as part of other software, including the GRAMMPS mission ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

.... for example, one based on RAPS [Firby87] The planners of some of these robot architectures run asynchronously with the control loop [Gat91 ] whereas the plmners of others run synchronously with the control loop [Bon97] Similarly, the plmners of some of these robot architectures run continuously [Sten95] [Lyons95] whereas the planners of others run only from time to time [Bon97] The planner of our robot architecture runs synchronously with the control loop and, depending on the navigation mode, either continuously (to control the robot in mode 3) or only from time to time (to plan the next ....

A. Stentz and M. Hebert, "A complete navigation system for goal acquisition in unknown environments", Autonomous Robots, vol. 2, No. 2, 1995, pp. 127-145.

....the efficiency of heuristic and incremental searches, yet still finds shortest paths. It achieves a speedup of one to two orders of magnitudes( over repeated A [10] searches by modifying previous search results locally. D has been extensively used on real robots, including outdoor HMMWVs [15], including UGV Demo II vehicles as part of the DARPA Unmanned Ground Vehicle program. It is currently also being integrated into Mars Rover prototypes and tactical mobile robot prototypes for urban reconnaissance [5, 9, 17] D is also used as part of other software, including the GRAMMPS mission ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

....operating on a flat plane with discrete obstacles[6] In recent years, there has been increased interest in algorithms capable of driving robots over rough terrain. Much of this work also maintains the assumption that space can be easily divided into what is and what is not traversable [1] 5][14]. Natural terrain rarely provides this simple distinction. There is generally a gradation between terrain that is easily Figure 1: Hyperion at the field experiment site. traversed and that which is completely impassable. Without modeling this variability, a robot will be paralyzed due to ....

A. Stentz and M. Hebert. "A Complete Navigation System for Goal Acquisition in Unknown Environments", Proc. IEEE/RSJ International Conference On Intelligent Robotic Systems, Aug 1995.

....these two locations, with the secondary requirement that the number of turns be minimized (shortest paths tend to hug obstacles, requiring many way points) A near optimal path might be slightly longer or have slightly more way points, but it must still be obstacle free. The A search algorithm [1] is a popular approach to generating optimally short paths, but it does not address the number of way points and we found it to be unacceptably slow for this application. We developed a depthfirst version of the A algorithm that finds a near optimal path orders of magnitude faster than the ....

....perception was one of the most important issues. Other well known robotic ground vehicles capable of autonomous navigation include the HERMIES series of robots developed at the Oak Ridge National Laboratory [3] and the NAVLAB series of Autonomous Land Vehicles at Carnegie Mellon University [4] [1]. Work on the Nomad rover at Carnegie Mellon and NASA Ames has touched on the need for path planning assistance in a telerobotic user interface, but no automation of obstacle detection or path planning is involved [5] Similarly for the Russian Marsokhod rover (also in collaboration with NASA ....

[Article contains additional citation context not shown here]

A. Stenz and M. Hebert, "A Complete Navigation System for Goal Acquisition in Unknown Environments," presented at ARPA Image Understanding Workshop, Monterey, CA, 1994.

....Executing it incurs cost c(l; o) 0 and reports the current location of the robot with certainty. We assume that it is possible to reach every location from every other location. 3 Example We illustrate the sensor planning problem using simple gridworlds, similar to those used on real robots [18]. Figure 1 (left) shows the gridworld that we use in our experiments, where the locations are squares. The start location is C1 and the goal location is J1. The robot can always sense its current location (O) or move north (N) east (E) south (S) or west (W) to an adjacent square. If the robot ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

....for the purpose of being sure to place footfalls in such a manner as to keep the robot supported for instance, by not stepping onto a rock which might roll, or a surface so steep that its feet are likely 27 to skid off. Another mapping system along these lines is that of Stentz and Hebert [47], for an outdoor vehicle (this one based on a HMMWV chassis, with a laser rangefinder mounted above the cab and computers in the back) In this case, the system is given the goal of navigating to a point at a particular range and bearing from its present location. That is, once again, the goal of ....

Anthony Stentz and Martial Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2:127--45, 1995.

....of the algorithm has been implemented on the Rocky7 prototype microrover. Much of the prior work in the area of rover navigation has concentrated on heuristic local trajectory generation, and particularly obstacle avoidance. This style of navigation has also been referred to as local navigation [79] and piloting [43] especially when paired with a goal seeking behaviour. Examples include: Gat s implementation of his ATLANTIS architecture (1991) 12] utilises a 10 heuristic planner based on Slack s navigation templates [75] for obstacle avoidance and goal seeking, as well as a ....

....the inability of D to plan some straight diagonal paths. In addition, the D planner has been implemented as part of a comprehensive system on a HMMWV, combined with a local navigator, SMARTY (similar to the Langer, Rosenblatt, and Hebert navigator described above) and the DAMN voting arbitrator [79]. D has been proven to be resolution complete in a bounded environment [76] Chatila and Lacroix 2D navigation algorithm (1995) 8] which has been implemented on the Adam prototype rover. From the sensed information, the That is, D is complete, assuming the resolution of its world ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2), August 1995.

.... can reach specified positions, using a behavior control approach [12, 7] Nomad [1] from Carnegie Mellon University operated more than 200 kilometers using various navigation modes (principally teleoperation) Another system from CMU, Navlab also performed autonomous navigation on natural terrain [17]. The adaptative navigation approach developed at LAAS within the framework of the EDEN experiment [3] demonstrated autonomous short range navigation in a natural environment gradually discovered by the robot. The approach combines various navigation modes (reflex, 2D and 3D) in order to adapt ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2), 1995.

....robots have operated with more autonomy than Dante II, but in much less demanding terrains and usually with close human oversight. As an example, the NavLab, with safety observer onboard, autonomously traveled several kilometers of moderate terrain in search of an obstacle free path to a goal (Stentz and Hebert 1995). Other robots have achieved long field missions, but required daily or more frequent human support (Thomas, Hine, and Garvey 1995; Wettergreen et al. 1999) The principal objective of the Dante project was to develop and demonstrate technologies which could lead to solutions for robotic ....

Stentz, A., and Hebert, M., 1995. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots 2(2).

....example, that it can move in each cell of a grid world only north, east, south, and west. The execution of some of these actions might not be possible because untraversable patches of terrain block the way, but the robot does not know where this happens since the terrain is unknown. The D method (Stentz 1995) then exhibits the following behavior: The robot repeatedly moves from its current location with minimal plan execution time to the goal location, assuming that unknown terrain is traversable. When it observes during plan execution that a particular patch of terrain is untraversable, it corrects ....

....with maximal local search spaces) but changes the heuristic function dynamically according to the assumption that unknown terrain is traversable. A backward search from the goal location can be used to update the heuristic function efficiently as untraversable patches of terrain get discovered (Stentz 1995). Figure 8 illustrates the difference between incremental best first search and D , assuming that the robot can observe the status of the four cells adjacent to it and can move to each of them, unless that cell is untraversable. D has been used on an autonomous high mobilitymultiwheeled vehicle ....

[Article contains additional citation context not shown here]

Stentz, A. and Hebert, M. 1995. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots 2(2):127--145.

....to the goal vertex or recognize that this is impossible. In the literature, this problem has been studied in the context of actual robot navigation problems. For example, the purpose of NAVLAB II, Carnegie Mellon s unmanned robot HMMWV, is to reach specified coordinates in unmapped static terrains [13]. To do so, the vehicle discretizes the unknown area into a coarse resolution map of square cells. Each cell is either traversable or untraversable. The vehicle always occupies exactly one cell and can move in all eight compass directions to traversable adjacent cells. Its sensors always detect ....

....vertex to the starting vertex, there is no traversable path from the starting vertex to the goal vertex either. Consequently, reaching the goal vertex is impossible in this case. Planning with the freespace assumption has been used on actual robots. For example, it has been applied to the outdoor [13] and indoor [8] navigation problems mentioned above. Planning with the freespace assumption has several advantages: It is easy to implement. Its computations can be done efficiently. The most time consuming step is to re calculate a shortest path after new knowledge about obstacles has been ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

....Executing it incurs cost c(l; o) 0 and reports the current location of the robot with certainty. We assume that it is possible to reach every location from every other location. 3 Example We illustrate the sensor planning problem using simple gridworlds, similar to those used on real robots [18]. Figure 1 (left) shows the gridworld that we use in our experiments, where the locations are squares. The start location is C1 and the goal location is J1. The robot can always sense its current location (O) or move north (N) east (E) south (S) or west (W) to an adjacent square. If the robot ....

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

....relevant time scale to the mission. 6.2. D : Dynamic Path Planning Dynamic path planning has been shown to be substantially faster than brute force replanning in dynamic environments [27] In addition, D has been demonstrated for single robot single goal tasks utilizing a similar architecture[29]. In this application, multiple instantiations of the D planning structure are used, one per goal in the mission statement, distributed based on the available computational capacity on each vehicle. This implies that the path to a goal being driven to by one robot may be planned on a processor ....

Stentz, A., Hebert, M. "A Complete Navigation System for Goal Acquisition in Unknown Environments," Autonomous Robots, 2(2), 1995.

....UGV. Data registration remains an interesting alternative but is not one we have invested in heavily. 4.3. Global Navigation Our global navigation system is based on the most recent version of the D (for dynamic A ) global planning algorithm originally developed on the Demo II program [3] 4][5]. The path planned is the lowest cost path available that moves the vehicle from its current position through a sequence of waypoint goals. The initial path is based on any available prior information. The path is recomputed continuously in order to preserve optimality by incorporating both new ....

Stentz, A., Hebert, M., "A Complete Navigation System for Goal Acquisition in Unknown Environments," Autonomous Robots, Vol. 2, No. 2, August 1995.

....reacting to sensory data as quickly as possible [4] 6] A more deliberative process, operating at a coarser resolution of information, is used to decide how best to direct the robot toward the goal. This approach has been used successfully in the past in several systems at Carnegie Mellon [2] 4][16]. Approaches to path planning for mobile robots can be broadly classified into two categories: those that use exact representations of the world (e.g. 7] 8] and those that use a discretized representation (e.g. 1] 7] 9] The main advantage of discretization is that the computational ....

....detection and avoidance. This obstacle map is fed to a global navigator running a path planning algorithm, such as FramedQuadtree D . Both the local and global navigators submit steering advice to an arbiter, which selects a steering command each time interval and passes it to the controller [16]. Fig. 15 shows the system modules and data flow. Fig. 14. The autonomous vehicle (HMMWV) used for our experiments. The vehicle is equipped with stereo vision, inertial guidance and GPS positioning. 17 Fig. 15. Data flow in the implemented system. Fig. 16 shows a successful traverse of the ....

A. Stentz and M. Hebert, A complete navigation system for goal acquisition in unknown environments, in: Autonomous Robots 2(2) (1995).

....Using a dual strategy of motionto goal and boundary following, the algorithm has successfully demonstrated path planning and execution aboard the JPL Rocky 7 Mars rover prototype. A CMU path planner [12] also produced with Mars rover navigation in mind, is based upon the D algorithm [14][15][16] Using a grid based approach, D uses sensor information to populate cells with traversability data, and plans paths that avoid hazards and that are distance optimal under current world knowledge. This scheme has been successfully demonstrated on a CMU ATRV robot [12] and will be ....

....autonomous navigation on the scale of 100 m, in accordance with near future Mars rover requirements. Neither addresses an anticipated requirement for autonomous travel on the scale of 10s or 100s of km. In the arena of grid search applied to path planning, the D (Dynamic A ) algorithm [14] [15], 16] was developed to balance the rigor of deliberative planning with the rapid response of reactive behavior. Like A , D operates on a map of cost values and finds the lowest cost path from the start to the goal. Employed on a vehicle in unknown terrain, D enables incremental changes to ....

A. Stentz, M. Hebert, A Complete Navigation System for Goal Acquisition in Unknown Environments , Autonomous Robots, Vol. 2, No. 2, August 1995.

....to avoid the obstacle could not be executed reliably or stably. Yet another common problem is the well known local minimum problem. It arises from the use of local rather than global optimization strategies. This problem is considered outside the scope of our work here, though one of the authors [Stentz, 1995] addresses this problem in our target environment in other writings. 2.3. Standard Architectural Model Consider the following hierarchical architectural model. This is a convenient view for organizing the description of the system. Spectrum of Characteristics. Higher levels of the hierarchy ....

Stentz, A. and Hebert, M. 1995. A Complete Navigation System for Goal Acquisition in Unknown Environments. Autonomous Robots, Vol. 2(2).

No context found.

A. Stentz and M. Hebert, "A Complete Navigation System for Goal Acquisition in Unknown Environments", Autonomous Robots, Vol. 2, No. 2, August 1995.

No context found.

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2), 1995.

No context found.

A. Stentz & M. Hebert. " A Complete Navigation System for Goal Acquisition in Unknown Environments", IEEE IROS, 1995.

No context found.

A. Stentz and M. Hebert, "A complete navigation system for goal acquisition in unknown environments," Autonomous Robots, Vol. 2, No. 2, 1995, pp. 127-145.

No context found.

A. Stentz and M. Hebert. A complete navigation system for goal acquisition in unknown environments. Autonomous Robots, 2(2):127--145, 1995.

No context found.

Stentz, A. and Hebert, M., "A Complete Navigation System for Goal Acquisition in Unknown Environments," Proc. of IROS '95, August 1995.

No context found.

A. Stentz, M. Hebert, A Complete Navigation System for Goal Acquisition in Unknown Environments, Carnegie-Mellon Technical Report CMU-RI-TR-94-7, 1994.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC