US20140074287A1

US20140074287A1 - Positive and negative obstacle avoidance system and method for a mobile robot

Info

Publication number: US20140074287A1
Application number: US14/001,266
Authority: US
Inventors: Matthew LaFary; George Paul
Original assignee: Adept Technology Inc
Current assignee: Omron Corp
Priority date: 2012-01-25
Filing date: 2013-01-25
Publication date: 2014-03-13
Anticipated expiration: 2033-01-25
Also published as: SG11201400966UA; EP3113055A1; EP2791842A4; PL2807526T3; US8688275B1; EP2791842B1; EP2791842A1; ES2812568T3; WO2013112842A1; US9592609B2; EP2807526A1; WO2013112907A1; US20140350725A1; EP2807526A4; DK2791842T3; SG11201400958XA; DK2807526T3; EP3113055B1; EP2807526B1

Abstract

Embodiments of the present invention provide methods and systems for ensuring that mobile robots are able to detect and avoid positive obstacles in a physical environment that are typically hard to detect because the obstacles do not exist in the same plane or planes as the mobile robot's horizontally-oriented obstacle detecting lasers. Embodiments of the present invention also help to ensure that mobile robots are able to detect and avoid driving into negative obstacles, such as gaps or holes in the floor, or a flight of stairs. Thus, the invention provides positive and negative obstacle avoidance systems for mobile robots.

Description

FIELD OF ART

This invention generally relates to mobile robots. More specifically, the invention is directed to systems and methods for detecting positive and negative obstacles in physical environments through which mobile robots move and preventing the mobile robots from driving into those obstacles.

BACKGROUND ART

Mobile robots, including autonomously-navigating mobile robots, inertially-guided robots, remote-controlled mobile robots, and robots guided by laser targeting, vision systems, roadmaps and beacons, to name a few examples, normally use horizontally-oriented laser sensors to scan the area in the mobile robot's direction of travel and to detect potential obstacles in the mobile robot's path. The horizontally-oriented lasers, which typically scan in two-dimensional planes roughly parallel with floor, work reasonably well for detecting objects that extend from the floor in a substantially perpendicular direction, so long as the obstacle intersects the horizontally-oriented scanning plane. However, they do not work well for detecting positive physical obstacles in the mobile robot's path that are parallel to the floor and/or obstacles that are not at the same height as the horizontally-oriented scanning plane. This means the mobile robots frequently have no way of detecting and avoiding positive obstacles, such as long tables with legs at the ends (and no legs in the middle), objects suspended from a ceiling or other structure, and obstacles that stick out from the edge of another object, like a keyboard tray. Mobile robots that use horizontally-oriented lasers for detecting obstacles also have problems detecting and avoiding unexpected negative obstacles, such as a hole in the floor, a descending flight of stairs, the end of a loading dock or the edge of a cliff.
Previous attempts to solve these problems have included, for example, attaching a plurality of vertically-oriented or randomly-oriented lasers to the mobile robots and using the vertical or randomly-oriented lasers to detect obstacles parallel to the floor, as well as holes or drop-offs in the floor. However, there have been a number of significant disadvantages associated with such solutions, including prohibitively-high cost associated with installing, using and maintaining a multiplicity of expensive lasers, and an unacceptably high number of false positives arising, for example, from gratings in the floor, which do not necessarily need to be avoided by the mobile robot.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide methods and systems for ensuring that mobile robots are able to detect and avoid positive obstacles in a physical environment that are typically hard to detect because the obstacles do not extend vertically from the floor. Embodiments of the present invention also help to ensure that mobile robots are able to detect and avoid driving into negative obstacles, such as gaps or holes in the floor, or a flight of stairs. In general, embodiments of the present invention include a positive obstacle avoidance system for use with mobile robots, a negative obstacle avoidance system for use with mobile robots, or both a positive and negative obstacle avoidance system operating in the same mobile robot. The inventive systems and methods work for a variety of different types of mobile robots (also known as “automated guided vehicles” or “AGVs”), including without limitation autonomously-navigating mobile robots, visually-guided robots, telepresence robots, haptic input-guided robots and laser- or beacon-following robots.
In one aspect of the invention, there is provided a positive obstacle avoidance system for use with a mobile robot in a physical environment, comprising a first laser attached to the mobile robot, a memory, a laser controller, a positive obstacle avoidance engine and a propulsion system. The first laser attached to the mobile robot is oriented to scan the physical environment in a first plane that is not parallel to the floor of the physical environment. The memory stores initial operating parameters and preferences for the positive obstacle avoidance engine, including: (1) a map defining a floor plan representing the physical environment and a two dimensional coordinate system delineating the floor plan, (2) a first data structure of two-dimensional coordinates from the floor plan, each two dimensional coordinate in the first data structure representing a location in the physical environment to be avoided by the mobile robot, (3) first laser placement information for the first laser, including the first laser's position on the mobile robot and the first laser's orientation (or tilt angle), and (4) a tolerance profile for the obstacle avoidance system. The tolerance profile typically includes a ceiling offset, a floor offset, a ceiling adjustment angle and a floor adjustment angle. Using a tolerance profile in the positive obstacle avoidance engine calculations reduces the number of false positives that might otherwise result if, for example, the floor or the ceiling is not exactly horizontal, or if the laser is not installed at exactly the right height or orientation relative to the floor, the ceiling or the direction of travel. The initial operating parameters may be received from a remote system via a wired or wireless communication interface on the mobile robot, stored in a database onboard the mobile robot, or hard-coded into the program instructions comprising the positive obstacle avoidance engine. The data structure used to store the two-dimensional coordinates representing locations in the physical environment that the mobile robot should avoid may comprise any suitable data structure for organizing and managing two-dimensional coordinate data, as would be known by those skilled in the computer arts, including without limitation, a collection of database records, a linked list, a table, an array, a tree, a heap, or a stack. In some embodiments and applications, the data structure of two-dimensional coordinates describes what is known in the art as an “occupancy grid” for the floor plan for the physical environment.
The laser controller receives a set of laser readings from the first laser, each laser reading corresponding to a location in the physical environment where the first laser detects a physical obstacle that may or may not need to be avoided by the mobile robot, depending, for example, on whether the object at the location is truly a physical object existing in the path of the mobile robot, or merely a spot on a distant part of the floor or the ceiling of the physical environment. Typically, the laser provides laser readings in polar coordinates (angle and distance to the detected obstacle), but other types of laser readings may be suitably used without departing from the scope of the invention. Thus, the laser readings received by the laser controller include a first plane angle (theta) and a first plane distance (radius=r) from a given site on the mobile robot.
For each reading in the set of laser readings received from the first laser, the positive obstacle avoidance engine, typically comprising computer software instructions executable on a microprocessor on board the mobile robot, determines whether the reading represents a spot on the floor or the ceiling of the physical environment, or otherwise represents an object existing between the floor and the ceiling that should be avoided by the mobile robot. If the reading identifies an obstacle, rather than a spot on the ceiling or the floor, the positive obstacle avoidance engine stores the x and y coordinates of the laser reading in the first data structure in order to “remember” that location, in accordance with the floor plan, as a location to be avoided. The positive obstacle avoidance engine accomplishes this by first converting the laser reading (theta and r) into a three-dimensional coordinate based on the first plane angle (theta), the first plane distance (r), the first laser position and said first laser tilt in the memory. The three-dimensional coordinate includes an x-component, a y-component and a z-component. Because the first laser may be attached to the mobile robot in a variety of different orientations, so long as its scanning plane is not horizontal (i.e., not parallel to the floor), the formula used to calculate the three-dimensional coordinate depends on the angle of the scanning plane relative to the floor and the direction of travel for the mobile robot. So, for example, when the first laser's scanning plane is perpendicular to the floor and parallel to the direction of the mobile robot's direction of travel, the positive obstacle avoidance engine may be programmed to convert the laser readings into a three-dimensional coordinate (x, y, z) using the formulas:
x-component=first laser x position+first plane distance*cosine(first plane angle),
y-component=first laser y position, and
z-component=first laser z position+first plane distance*sin(first plane angle).
Next, the positive obstacle avoidance engine determines a minimum ceiling height and a maximum floor height for the three-dimensional coordinate based on the x-component, the y-component and the tolerance profile. The maximum floor height may be calculated, for example, by using the formula:
maximum floor height=floor offset+A*sin(floor adjustment angle),
wherein, A=the distance from the current position to the coordinate (x-component, y-component) in the two-dimensional coordinate system. The minimum ceiling height may be calculated according to the formula:
minimum ceiling height=ceiling offset+A*sin(ceiling adjustment angle),
wherein, A=the distance from the current position to the coordinate (x-component, y-component) in the two-dimensional coordinate system.
If the z-component is between the minimum ceiling height and the maximum floor height, then the positive obstacle avoidance engine is considered to have detected a positive obstacle suspended above the floor and below the ceiling. This obstacle could be the horizontal part of a table, a keyboard tray sticking out of a desk or workstation, or some other cantilevered object. In this case, the positive obstacle avoidance engine will add the x-component and the y-component of the three-dimensional coordinate (representing the two-dimensional locations on the floor plan beneath the table, keyboard tray or other object) to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot. In this fashion, the positive obstacle avoidance engine builds a data structure of two-dimensional coordinates (or adds new two-dimensional coordinates to a previously-existing data structure) that identifies all of the locations on the floor plan where a positive obstacle has been found.
The propulsion system typically comprises a collection of hardware and software components that cause the mobile robot to drive about the physical environment under control of the first data structure. This means the propulsion system prevents the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure by plotting a path around the locations, if possible, or bringing the mobile robot to a stop if no path around the obstacle is available. In some embodiments, the mobile robot receives a command from another computer system to move the mobile robot from its current position to a specified location in the physical environment. In such cases, the mobile robot may include a wired or wireless communication interface to enable receiving such commands, and the propulsion system may encompass a path planning engine that calculates, in accordance with the floor plan, a path between the current position and the specified location, wherein the path is calculated so as to avoid the locations in the physical environment represented by the first data structure of two-dimensional coordinates. Then the propulsion system, which may include navigation and locomotion components (e.g., motors and wheels and a microcontroller to operate the motors and wheels) automatically drives the mobile robot along the calculated path, thereby avoiding the obstacle.
Multiple non-horizontal lasers oriented to scan the physical environment in non-horizontal planes may be used to enhance the mobile robot's ability to detect and avoid positive obstacles that do not extend vertically from the floor or intersect the mobile robot's horizontal scanning planes, and are therefore undetected by the mobile robot's horizontal lasers. Thus, in addition to the first non-horizontal laser scanning in the first plane, a second non-horizontal laser may be attached to the mobile robot and oriented to scan the physical environment in a second plane that is also not parallel to the floor of the physical environment. In such embodiments, the memory stores a second set of initial operating parameters and preferences for use by the positive obstacle avoidance engine in processing the laser readings from the second non-horizontal laser, including: (1) a second data structure of two-dimensional coordinates from the floor plan, each two dimensional coordinate in the second data structure representing another location in the physical environment to be avoided by the mobile robot, and (2) placement information for the second laser, including a second laser position and a second laser tilt. In this case, the laser controller receives a second set of laser readings from the second laser, each laser reading corresponding to a location in the physical environment where the second laser detects a physical obstacle in the second plane. The positive obstacle avoidance engine performs the same calculations for each laser reading received from the second set of laser readings in order to determine whether the detected obstacle represents the expected floor or ceiling, based on distance and angle from the robot, or otherwise represents some obstacle located between the floor and the ceiling in the physical environment that needs to be avoided. Thus, the positive obstacle avoidance engine converts each laser reading into a second three-dimensional coordinate based on the second plane angle (theta), the second plane distance (r), the second laser position and the second laser tilt. The positive obstacle avoidance engine then determines an minimum allowable ceiling height and a maximum allowable floor height for the second three-dimensional coordinate based on the x-component of the second three-dimensional coordinate, the y-component of the second three-dimensional coordinate and the tolerance profile. Then the engine determines whether the second z-component lies between the allowable floor height and the allowable ceiling height based on the z-component's distance from the robot. If so, the positive obstacle avoidance engine adds the x-component and the y-component of the second three-dimensional coordinate to the second data structure of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot. In this manner, the positive obstacle avoidance engine builds and populates a second collection or list of two-dimensional coordinates from the floor plan that the mobile robot should avoid while driving about the physical environment. The propulsion system drives the mobile robot about the physical environment under control or influence of the first data structure and the second data structure, while preventing the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in both the first data structure and the second data structure. It will be understood by those skilled in the art that embodiments of the present invention may in fact store all of the two-dimensional coordinates for both the first laser and the second laser in a single data structure, or a multiplicity of different data structures, instead of two discreet data structures, without departing from the scope of the claimed invention. Any number of data structures may be used, so long as the propulsion system accesses the stored two-dimensional coordinate data to determine which parts of the floor plan should not be driven into by the mobile robot.
In another aspect, embodiments of the present invention also provide a negative obstacle avoidance system for use with a mobile robot in a physical environment. The negative obstacle avoidance system operates to prevent the mobile robot from driving into a hole or ditch, driving off the edge of a cliff, or driving across an unacceptably large gap in the floor. In essence, the negative obstacle avoidance system does this by comparing currently-received floor distance readings from one or more non-horizontally-oriented lasers with a tolerance profile, which defines, among other things, the largest gap in the floor (i.e., a “gap profile”) that the mobile robot will be permitted to traverse. Embodiments of the present invention, however, also use the tolerance profile settings to avoid recognizing too many false positives in the data, which could arise, for example, when the mobile robot is approaching a relatively safe grating, small bump or small gap in the floor. The tolerance profile settings also reduce the number of false positives that arise when the laser data readings are less than perfect. This might occur, for example, when the laser light coming from the mobile robot strikes something too reflective and at an angle that prevents the light from returning to the sensor, or when the laser light coming from the mobile robot strikes something that is not reflective enough, and therefore absorbs too much of the light for any of it to return to the mobile robot.
In general, the negative obstacle avoidance system comprises a first laser attached to the mobile robot, a memory, a laser controller, a negative obstacle avoidance engine and a propulsion system. The first laser is oriented to scan the physical environment in a first plane that is not parallel to the floor of the physical environment. The memory stores a set of initial operating parameters and preferences for use by the negative obstacle avoidance engine, including: (1) a map defining a floor plan representing the physical environment and a two dimensional coordinate system delineating the floor plan, (2) a first data structure of two-dimensional coordinates from the floor plan, each two dimensional coordinate in the first data structure representing locations in the physical environment to be avoided by the mobile robot, (3) first laser placement information for the first laser, including a first laser position and a first laser tilt, and (4) a tolerance profile comprising a maximum floor offset, a minimum floor offset, a maximum floor adjustment angle, a minimum floor adjustment angle and a maximum allowable floor gap.
The laser controller receives a set of laser readings from said first laser, each laser reading corresponding to a location in the physical environment within said first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot. The negative obstacle avoidance engine first converts each laser reading into a three-dimensional coordinate based on the first plane angle, the first plane distance, the first laser position and the first laser tilt. Next, the negative obstacle avoidance engine determines a maximum allowable floor height and a minimum allowable floor height for the three-dimensional coordinate based on the x-component, the y-component and the tolerance profile. The engine then determines if the z-component falls between the maximum allowable floor height and the minimum allowable floor height. If it does, the negative obstacle avoidance system changes an indicator, such as a flag in the memory to designate the x-component and the y-component of the three-dimensional coordinate as the coordinate corresponding to the last good floor reading. If the z-component does not fall between the maximum allowable floor height and the minimum allowable floor height, then the negative obstacle avoidance engine determines whether the distance G between the last good floor reading coordinate and the location in the two-dimensional coordinate system defined by the x-component and the y-component exceeds the maximum allowable floor gap stored in the memory. If the distance G exceeds the maximum allowable floor gap, the negative obstacle avoidance engine adds the x-component, the y-component and the last good floor reading coordinates to the first data structure of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot. The propulsion system then drives the mobile robot in the physical environment under control of the first data structure, so as to prevent the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure. Thus, the mobile robot, operating under the control of the propulsion system, which itself operates under the influence of the first data structure of two-dimensional coordinates identifying all of the two-dimensional coordinates in the floor plan to be avoided, avoids driving into the gap, ditch or hole, or off of a cliff.
In yet another aspect there is provided a method for avoiding positive obstacles in a mobile robot comprising a memory and a propulsion system. The first step in the method comprises storing the map, the first data structure of two-dimensional coordinates, the first laser placement information and the tolerance profile in the memory. In the next step, a set of laser readings from the first laser are received, each laser reading corresponding to a location in the physical environment where said first laser detects a physical obstacle in the first plane, and each laser reading having a first plane angle (theta) and a first plane distance (radius=r) from a given spot on the mobile robot. Next, each laser reading in the set of laser readings is automatically processed by the mobile robot by: (1) converting each laser reading into a three-dimensional coordinate based on the first plane angle (theta), said first plane distance (r), the first laser position and the first laser tilt. The three-dimensional coordinate includes an x-component, a y-component and a z-component. Then, a minimum ceiling height and a maximum floor height for the three-dimensional coordinates are determined based on the x-component, the y-component and the tolerance profile. If the z-component falls between the minimum ceiling height and the maximum floor height, then the laser reading is considered to have detected an obstacle that (such as a table surface or overhang) that should be avoided, and the x-component and the y-component of the three-dimensional coordinate (which point to the spot on the floor below the table or overhang) are added to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot. In the final step, the propulsion system drives the mobile robot in the physical environment under the influence or control of the first data structure. As previously stated, this means the propulsion system periodically checks the data in the first data structure of two-dimensional coordinates and prevents the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure. By avoiding the spots on the floor below the table or overhang, the mobile robot also avoids driving into the table or overhang. In some embodiments, the mobile robot does its own path planning and navigation. In these embodiments, the method further includes the steps of: (1) receiving on the mobile robot a command to move from a current position to a specified location in the physical environment, (2) calculating on the mobile robot, in accordance with the floor plan, a path between the mobile robot's current position and the specified location that avoids the locations in the physical environment represented by the coordinates in the first data structure of two-dimensional coordinates, and (3) activating the propulsion system to automatically drive the mobile robot along the path.
In still another aspect, the present invention provides a method for avoiding negative obstacles in a physical environment using a mobile robot comprising a memory and a propulsion system. In this aspect, the method begins by storing in the memory a map defining the floor plan representing the physical environment and a two dimensional coordinate system delineating the floor plan, a first data structure of two-dimensional coordinates from the floor plan, the two dimensional coordinates in the first data structure representing locations in the physical environment to be avoided by the mobile robot, first laser placement information for a first laser attached to the mobile robot, the first laser being oriented to scan the physical environment in a first plane that is not parallel to the floor of the physical environment. The first laser placement information includes a first laser position and a first laser tilt, and a tolerance profile comprising a maximum floor offset, a minimum floor offset, a maximum floor adjustment angle, a minimum floor adjustment angle and a maximum allowable floor gap. The next step comprises receiving a set of laser readings from the first laser, each laser reading corresponding to a location in the physical environment within the first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot. Then, each laser reading in the set of laser readings is converted into a three-dimensional coordinate based on the first plane angle (theta), said first plane distance (r), the first laser position and the first laser tilt. The three-dimensional coordinate has an x-component, a y-component and a z-component. In the next step, the mobile robot determines a maximum allowable floor height and a minimum allowable floor height for the three-dimensional coordinate based on the x-component, the y-component and the tolerance profile. If the z-component falls between the maximum allowable floor height and the minimum allowable floor height, the mobile robot changes an indicator, such as flag in the memory to designate the x-component and the y-component of the three-dimensional coordinate as the last good floor reading. However, if the z-component does not fall between the maximum allowable floor height and the minimum allowable floor height, the mobile robot next determines whether the distance G between the last good floor reading and the location in the two-dimensional coordinate system defined by the x-component and the y-component exceeds the maximum allowable floor gap. If it does, then the system has found a gap in the floor that the mobile robot cannot safely traverse, in which case the system adds the x-component, the y-component and the last good floor reading to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot. In the final step of the method, the system causes the propulsion system to drive the mobile robot in the physical environment under the influence or control of the first data structure, the propulsion system preventing the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention and various aspects, features and advantages thereof are explained in more detail below with reference to exemplary and therefore non-limiting embodiments and with the aid of the drawings, which constitute a part of this specification and include depictions of the exemplary embodiments. In these drawings:

FIGS. 1A and 1B together illustrate, by way of example, one of the disadvantages of conventional laser-based obstacle avoidance systems for mobile robots.

FIG. 2 illustrates, by way of non-limiting example, the placement and orientation of an additional scanning laser in accordance with an exemplary embodiment of the present invention.

FIG. 3 contains a schematic diagram that illustrates, by way of non-limiting example, how a scanning laser attached to a mobile robot, and oriented to operate in accordance with exemplary embodiments of a positive obstacle avoidance system of the present invention, detects horizontally-oriented objects in the physical environment where the mobile robot operates.

FIG. 4 contains a schematic diagram illustrating, by way of example, how a scanning laser attached to a mobile robot, and oriented to operate in accordance with embodiments of the present invention, detects an unacceptably large gap in the floor of the physical environment.

FIG. 5 shows a high-level block diagram illustrating the major physical and logical components of a mobile robot with a positive and negative obstacle avoidance system in accordance with exemplary embodiments of the present invention.

FIG. 6 shows, in printed form, an example of the coordinate data in the data structure of two-dimensional coordinates in some embodiments of the present invention.

FIG. 7 shows a high-level flow diagram illustrating steps that may be performed by a positive and negative obstacle avoidance system, like the one shown in FIG. 5, operating in accordance with embodiments of the present invention to determine the locations of positive obstacles in the physical environment.

FIG. 8 shows a high-level flow diagram illustrating steps that may be performed by a positive and negative obstacle avoidance system, like the one shown in FIG. 5, operating in accordance with embodiments of the present invention to determine the locations of negative obstacles in the physical environment.

FIG. 9 illustrates, by way of example only, a known algorithm a mobile robot might use to plan a path between the mobile robot's current position and a particular location in accordance with a map.

FIG. 10 illustrates, by way of example, some of the data content of a map file defining a floor plan for the physical environment. The map file is stored in the memory of the mobile robot according to embodiments of the invention.

FIG. 11 shows a graphical representation of the map file illustrated in FIG. 10.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

FIGS. 1A and 1B together illustrate, by way of non-limiting example, one of the disadvantages of conventional laser-based obstacle avoidance systems for mobile robots. As shown in FIGS. 1A and 1B, mobile robot 10 is equipped with a horizontally-oriented scanning laser (not shown), which emits laser light rays 15 that fan out in front of the robot in a substantially horizontal plane that is roughly parallel to the floor or other surface upon which the mobile robot travels. When mobile robot 10 approaches an obstacle that is resting on the floor of the physical environment, or very near the floor, such as obstacle 20 in FIG. 1A, the laser light rays 15 strike the obstacle 20 and are reflected back to the mobile robot 10. Sensors in the laser detect and process the reflections, thereby informing the mobile robot 10 that there is an obstacle 20 in its path that needs to be avoided. Thus, the obstacle avoidance and locomotion controllers on board mobile robot 10, if any, cause mobile robot 10 to go around obstacle 20 or otherwise brake to come to a stop before mobile robot 10 can collide with the obstacle 20. This process is sometimes called “laser localization,” or “Monte Carlo localization with a laser.” However, as shown in FIG. 1B, when mobile robot 10 approaches an obstacle 25 suspended above the floor or travel surface in such a way that obstacle 25 does not lie in the same plane as the laser light rays 15 emanating from the horizontally-oriented laser on mobile robot 10, then the laser light rays 15 will not strike and thus will not be reflected back from obstacle 25. Since mobile robot 10 does not detect light bouncing back at it from the distance corresponding to the location of obstacle 25, it does not “know” that there is an obstacle 25 in its path and, consequently, collides with obstacle 25, potentially causing severe damage to mobile robot 10, obstacle 25, cargo 12, or all of them.
Although it would be possible to detect obstacle 25 by raising the scanning laser to a higher position on mobile robot 10, this solution would only lead to mobile robot 10 not being able to detect and avoid colliding with obstacles closer to the floor, such as obstacle 20 shown in FIG. 1A. Depending on the height of the mobile robot, as well as the physical environment where the mobile robot operates, adding a multiplicity of horizontally-oriented scanning lasers to mobile robot 10 in order to account for every possible obstacle may be impractical due to the expense and the trouble of purchasing, installing and maintaining a large number of lasers for every mobile robot in an organization, as well as managing, crunching and using all of the data that such a multiplicity of installed lasers would produce.
FIG. 2 illustrates, by way of non-limiting example, the placement and orientation of an additional scanning laser in accordance with an exemplary embodiment of the present invention. As shown in FIG. 2, scanning laser 27 is attached to mobile robot 10 and oriented so that it emits laser light rays 30 in a plane that is not parallel to the floor of the physical environment. In this orientation, scanning laser 27 emits laser light rays 30 in a plane that is much more likely to intersect with and be reflected from any obstacle, including obstacle 35, in mobile robot 10's path. Thus laser light rays 30 are much more likely to be reflected back to mobile robot 10 when mobile robot 10 approaches tables with legs at the ends, protruding keyboard trays, obstacles suspended from the ceiling or suspended from other objects, and the like. Although the scanning laser 27 may be installed so that the laser light rays 30 emanating from scanning laser 27 fall into a plane that is substantially perpendicular to the floor, scanning the physical environment in a plane that is perpendicular to floor is not required. Orienting the lasers to scan other planes will work just as well or better, depending on the lasers used and the physical environment, so long as the lasers are installed so that they scan in planes that are not parallel to the floor, or more accurately, parallel with the objects that are most likely to form obstacles that the mobile robot needs to avoid, such as tables and keyboard trays.
FIG. 3 contains a schematic diagram that illustrates, by way of non-limiting example, how a scanning laser 315 attached to a mobile robot 310, and oriented to operate in accordance with exemplary embodiments of a positive obstacle avoidance system of the present invention, detects horizontally-oriented objects in mobile robot 10's path of travel. As shown in FIG. 3, scanning laser 315 produces laser light rays that intersect with a horizontally-oriented obstacle 320 in mobile robot 310's path of travel. As illustrated by FIG. 3, a multiplicity of laser light rays may strike and be reflected back toward mobile robot 310, depending on the relative heights of scanning laser 315 and obstacle 320. It should be understood that, even if scanning laser 315 and obstacle 320 are at exactly the same height, the non-horizontal orientation of the plane of laser light produced by scanning laser 315 ensures that at least some of the laser light rays will strike the obstacle 320. Thus, while mobile robot 310 may not be capable of “seeing” multiple spots on the top surface of obstacle 320 (as shown in FIG. 3), because the obstacle 320 is at the same height as the scanning laser 315, it will still “see” the nearest edge of the horizontally-oriented obstacle as the mobile robot 310 approaches the obstacle 320, and therefore take steps to avoid it, as will be described in more detail below.
In accordance with embodiments of the present invention, FIG. 3 also illustrates the locations of the minimum allowable ceiling height 330 and maximum allowable floor height 335 calculated and used by the positive obstacle avoidance engine, as described herein. The slope in the minimum allowed ceiling height 330 and the maximum allowable floor height 335, as well as their offsets from the real ceiling 331 and real floor 332, respectively, permits mobile robots operating according to embodiments of the invention to function properly even though the ceiling or the floor, or both of them, may not be exactly horizontal. Thus, the mobile robot 310 will not determine, erroneously, that the floor or the ceiling is an obstacle in its path that must be avoided merely because the floor happens to rise a little, or because the ceiling happens to fall a little, as the distance from the robot increases, so long as the laser readings received by the mobile robot 310 indicate that the object detected is at or below the maximum allowable floor height 335, or at or above the minimum allowable ceiling height 330.
FIG. 4 contains a schematic diagram illustrating, by way of example, how a scanning laser 415 attached to a mobile robot 410, and oriented to operate in accordance with embodiments of the present invention, detects an unacceptably large gap 420 in the floor 405 of the physical environment. As shown in FIG. 4, scanning laser 415 is attached to mobile robot 410 and oriented to scan the physical environment in a plane that intersects the floor 405. As explained herein, mobile robot 410 has a negative obstacle avoidance engine that is configured to remember the last good floor reading obtained by the engine and constantly compares that last good floor reading with current readings in order to determine if it is safe for the mobile robot 410 to traverse the gap 420. If the gaps are small, for example, such as would be the case when the gap is one of the holes in a screen or grate, then embodiments of the present invention are typically configured to permit the mobile robot to continue driving over the gaps in the screen or grate because such screens or grates will not pose a risk or danger to the mobile robot 410.
FIG. 5 shows a high-level block diagram illustrating the major physical and logical components of a mobile robot 501 having a positive and negative obstacle avoidance system in accordance with exemplary embodiments of the present invention. As shown in FIG. 5, the mobile robot 501 includes a non-horizontal laser (or “first laser”) 505, a non-horizontal laser controller 510, a positive and negative obstacle avoidance engine 515, a memory 503 and a propulsion system 550. The non-horizontal laser 505 on the mobile robot 501 is oriented to scan the physical environment in a first plane (not shown) that is not parallel to the floor of the physical environment. The non-horizontal laser is programmatically coupled to a non-horizontal laser controller 510, that receives the non-horizontal laser's readings are passes that information along to the positive and negative obstacle avoidance engine 515. Suitable lasers and laser controllers to use for these purposes may be obtained, for example, from Hokuyo Urg, of Japan (Part Nos. URG-04LX and URG-04LX-UG01).
Typically, but not necessarily, mobile robots operating according embodiments of the present invention will also have one or more other range devices 580, including another scanning laser, that the mobile robot 501 uses, for example, to scan in a horizontal plane that is parallel to the floor. The horizontally-oriented laser and other range devices 580 send readings to one or more range device controllers 585, which uses the data to populate the data structure of two-dimensional coordinates 520 with two-dimensional (x,y) coordinates representing locations in the physical environment that the mobile robot 501 should avoid. The other range devices 580 may be very useful, for example, for detecting vertically-oriented objects extending from the floor of the physical environment, which may not be easy to detect with the non-horizontal laser 505, especially if the non-horizontal laser 505 happens to be oriented to scan in a plane that is perpendicular to floor. Notably, although shown as separate controllers in FIG. 5, it will be recognized by those skilled in the art, upon reading this disclosure, that embodiments of the present invention may use a single range device controller to receive readings from both the horizontally-oriented lasers and the non-horizontally-oriented laser.
The memory 503 stores initial operating parameters and preferences for the positive obstacle avoidance engine 515, including a map 530 defining a floor plan 535 representing the physical environment and a two dimensional coordinate system delineating the floor plan 535. Exemplary data content for a computer file comprising the map 530 is shown in FIG. 10. FIG. 11 shows a graphical representation of the map 530. The initial operating parameters stored in the memory 503 also include a data structure of two dimensional coordinates 520. Each two dimensional coordinate in the data structure 520 represents a location in the physical environment to be avoided by the mobile robot 501 because some obstacle has been detected at that location. In some embodiments, coordinates for virtual obstacles 525, including things like forbidden zones, need to enter zones, single robot zones, robot avoidance zones, one-way zones, and the like, may be defined by the system operator and added to the data structure of two-dimensional coordinates to control where mobile robots can and cannot go in the physical environment, even though no physical obstacles exist in those locations. FIG. 6 shows, in printed form, an example of the coordinate data stored in the data structure of two-dimensional coordinates 520 in some embodiments of the present invention.
Returning to FIG. 5, the memory 503 also includes laser placement information 540 for the non-horizontal laser 505, including the non-horizontal laser 505's position on the mobile robot 501 and the first laser's tilt angle (both values being relative to the center of the robot, for example). The memory 503 also holds a tolerance profile 545 for use with the positive and negative obstacle avoidance engine 515. As previously stated, the tolerance profile 545 typically includes a ceiling offset, a floor offset, a ceiling adjustment angle and a floor adjustment angle to be used by the positive and negative obstacle avoidance engine 515 to avoid generating too many false positives caused by sloping ceilings or floors, or variations in the position and tilt of the installed laser.
The propulsion system 550 may comprise a combination of hardware, such as motors and wheels 570, and software processors and/or controllers, such as path planning engine 555 and locomotion controller 560, that when executed by a microprocessor on board the mobile robot 501 (the microprocessor is not shown), cause the mobile robot 501 to avoid driving into the locations in the physical environment represented by the coordinates in the data structure of two dimensional coordinates 520. In other words, the propulsion system 550 is typically configured to periodically check the coordinates in the data structure 520 in order to ensure, for example, that no coordinates from the data structure 520 will be used in a path planned by the path planning engine 565. In some embodiments, avoiding the locations in the physical environment represented by the coordinates in the data structure 520 will mean calculating paths around the prohibited locations. In other embodiments, the mobile robot 501 may slow down and/or come to a complete stop if the path to the current destination is blocked by a prohibited location or because proceeding along the intended path would cause the mobile robot 510 to drive into a gap in the floor.
As shown in FIG. 5, the propulsion system 550 may optionally include a path planning engine 565, which determines, in accordance with the map 530 and the floor plan 535 for the physical environment, an optimum path the mobile robot 501 should take to travel from its current position in the physical environment to a specified location. A variety of different path planning techniques are known and used by those skilled in the art to accomplish such path planning for mobile robots. FIG. 9, discussed below, shows a high-level flow diagram illustrating, by way of example, the steps that might be performed by one such path planning engine encompassed by the propulsion system 550. Another path planning technique is discussed in detail in Chapter 7 of the book “Artificial Intelligence and Mobile Robots,” First Edition, published in 1998 by AAAI Press, and edited by David Kortenkamp, R. Peter Bonnaso and Robin Murphy. For purposes of the present invention, the path planning engine 555 plans a path that avoids all of the locations that have been identified by the positive and negative obstacle avoidance engine 515 as containing obstacles.
FIG. 7 shows a high-level flow diagram illustrating steps that may be performed by a positive and negative obstacle avoidance system, like the one shown in FIG. 5, operating in accordance with embodiments of the present invention to determine the locations of positive obstacles in the physical environment. As shown in FIG. 7, the first step, step 705, comprises receiving angle and distance information (theta and r) from the non-horizontal laser 505 and non-horizontal laser controller 510. Step 705 in FIG. 7 also includes initializing certain operating parameters for the positive and negative obstacle avoidance engine 515. The remaining steps 710 through 725 are performed by the positive and negative obstacle avoidance engine 515 to populate the data structure of two-dimensional coordinates 520 with two-dimensional coordinates representing locations in the physical environment that need to be avoided.
More specifically, in step 710, the positive and negative obstacle avoidance engine 515 converts the laser reading (theta and r) obtained in step 705 into a three-dimensional coordinate (x, y, z) based on the first plane angle (theta), the first plane distance (r), the first laser position and said first laser tilt stored in the memory 503 at step 705. The three-dimensional coordinate includes an x-component, a y-component and a z-component. As discussed above, the first laser may be attached to the mobile robot in a variety of different orientations. But when the first laser's scanning plane is perpendicular to the floor and parallel to the direction of the mobile robot's direction of travel, the positive obstacle avoidance engine is programmed to convert the laser readings into a three-dimensional coordinate (x, y, z) using the formulas:
x-component=first laser x position+first plane distance*cosine(first plane angle),
y-component=first laser y position, and
z-component=first laser z position+first plane distance*sin(first plane angle).
Next, at step 715, the positive obstacle avoidance engine 515 determines a maximum floor height and a minimum ceiling height for the three-dimensional coordinate based on the x-component, the y-component and the tolerance profile 545 stored in the memory 503. As previously stated, the maximum floor height may be calculated according to the formula:
maximum floor height=floor offset+A*sin(floor adjustment angle),
where A=the distance from the current position to the coordinate (x-component, y-component) in the two-dimensional coordinate system. The minimum ceiling height may be calculated according to the formula:
minimum ceiling height=ceiling offset+A*sin(ceiling adjustment angle),
where A=the distance from the current position to the coordinate (x-component, y-component) in the two-dimensional coordinate system.
Next, at step 720, the positive and negative obstacle avoidance engine 515 determines whether the z-component is between the minimum ceiling height and the maximum floor height. If it is, then the positive and negative obstacle avoidance engine construes this as an indication that the location is occupied by a positive obstacle suspended above the floor, such as the horizontal component of a table, a keyboard tray extending out of a desk or workstation, or possibly some other cantilevered object. Accordingly, at step 725, the positive obstacle avoidance engine 515 will then add the x-component and the y-component of the three-dimensional coordinate (representing the spot on the floor below the table, keyboard tray or other object) to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot. If there are more distance and angle readings from the non-horizontal laser, then execution returns to step 710, upon which the next three-dimensional coordinate is processed. In this fashion, the positive obstacle avoidance engine builds the data structure of two-dimensional coordinates (or adds new two-dimensional coordinates to a previously-existing data structure) that identifies all of the locations on the floor plan where a positive obstacle has been found. Notably, the data structure of two dimensional coordinates built by the positive and negative obstacle avoidance system may also contain coordinates for obstacles that were found, or could also be found, by a horizontally-oriented laser on the mobile robot. In other words, the positive and negative obstacle avoidance system may be configured to operate completely independently from the operation of other obstacle avoidance systems tied, for example, to a horizontally-oriented later on the mobile robot.
FIG. 8 shows a high-level flow diagram illustrating steps that may be performed by a positive and negative obstacle avoidance system, like the one shown in FIG. 5, operating in accordance with embodiments of the present invention to determine the locations of negative obstacles in the physical environment. First, in step 805, laser readings are received from the non-horizontal laser 505 and the non-horizontal laser controller 510. Also in this step, a set of initial operating parameters and preferences for use by the positive and negative obstacle avoidance engine 515 is stored in the memory, including laser placement information for the non-horizontal laser, including a first laser position and a first laser tilt, and a tolerance profile 545 comprising a maximum floor offset, a minimum floor offset, a maximum floor adjustment angle, a minimum floor adjustment angle and a maximum allowable floor gap.
Then, in step 810, the positive and negative obstacle avoidance engine converts each laser reading into a three-dimensional coordinate based on the first plane angle, the first plane distance, the first laser position and the first laser tilt. Next, at step 815, the negative obstacle avoidance engine determines a maximum allowable floor height and a minimum allowable floor height for the three-dimensional coordinate based on the x-component, the y-component and the tolerance profile, and determines in step 820 whether the z-component falls between the maximum allowable floor height and the minimum allowable floor height. If it does, the positive and negative obstacle avoidance engine 515 changes an indicator, such as a flag, in the memory to designate the x-component and the y-component of the three-dimensional coordinate as the coordinate corresponding to the last good floor reading (step 825). On the other hand, if the z-component does not fall between the maximum allowable floor height and the minimum allowable floor height, then, in steps 830 and 835, the positive and negative obstacle avoidance engine 515 determines whether the distance G between the last good floor reading coordinate and the location in the two-dimensional coordinate system defined by the x-component and the y-component exceeds the maximum allowable floor gap stored in the memory 503. If the distance G exceeds the maximum allowable floor gap, the positive and negative obstacle avoidance engine 515 adds the x-component, the y-component and the last good floor reading coordinate to the data structure 520 of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot 501 (step 840). When all of the non-horizontal laser readings are processed, positive and negative obstacle avoidance engine will have populated the data structure 520 with a list of coordinates that need to be avoided. This information may then be used by a path planning engine, an autonavigation engine, or both.
FIG. 9 illustrates, by way of example only, a known algorithm a mobile robot might use to plan a path between the mobile robot's current position and a particular location in accordance with a map and a floor plan delineated by a two-dimensional coordinate system. In general, the steps of the path planning algorithm include receiving the job location (step 905), planning a global path to the job location (step 910), planning a local path around detected and remembered obstacles (step 915), finding an allowable set of movement commands for the mobile robot in order to follow the path (step 925) and sending the movement commands to a locomotion controller or motors (step 930).
Although the exemplary embodiments, uses and advantages of the invention have been disclosed above with a certain degree of particularity, it will be apparent to those skilled in the art upon consideration of this specification and practice of the invention as disclosed herein that alterations and modifications can be made without departing from the spirit or the scope of the invention, which are intended to be limited only by the following claims and equivalents thereof.

Claims

1. In a mobile robot comprising a memory and a propulsion system, a method for avoiding positive obstacles in a physical environment, the method comprising:

a) storing in the memory

(i) a map defining a floor plan representing the physical environment and a two dimensional coordinate system delineating the floor plan,

(ii) a first data structure of two-dimensional coordinates from the floor plan, the two-dimensional coordinates in the first data structure representing locations in the physical environment to be avoided by the mobile robot,

(iii) first laser placement information for a first laser attached to the mobile robot, said first laser being oriented to scan the physical environment in a first plane that is not parallel to the floor of the physical environment, wherein said first laser placement information includes a first laser position and a first laser tilt, and

(iv) a tolerance profile comprising a ceiling offset, a floor offset, a ceiling adjustment angle and a floor adjustment angle,

b) receiving a set of laser readings from said first laser, each laser reading corresponding to a location in the physical environment where said first laser detects a physical obstacle in said first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot;

c) for each laser reading in said set of laser readings

(i) converting said each laser reading into a three-dimensional coordinate based on said first plane angle, said first plane distance, said first laser position and said first laser tilt, wherein said three-dimensional coordinate comprises an x-component, a y-component and a z-component,

(ii) determining a minimum ceiling height and a maximum floor height for the three dimensional coordinate based on the x-component, they-component and the tolerance profile, and

(iii) if the z-component is between the minimum ceiling height and the maximum floor height, adding the x-component and they-component to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot; and

d) causing the propulsion system to drive the mobile robot in the physical environment under control of the first data structure, wherein the propulsion system prevents the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure.

2. The method of claim 1, further comprising:

a) receiving on the mobile robot a command to move from a current position to a specified location in the physical environment;

b) calculating on the mobile robot, in accordance with the floor plan, a path between said current position and said specified location, said path avoiding the locations in the physical environment represented by the first data structure of two-dimensional coordinates; and

c) activating the propulsion system to automatically drive the mobile robot along the path.

3. The method of claim 1, wherein said first plane is perpendicular to the floor of the physical environment.

4. The method of claim 1, wherein said maximum floor height is calculated by using the formula:

maximum floor height=floor offset+A*sin(floor adjustment angle),

wherein, A=the distance from the current position to the coordinate (x-component, y-component) in the two-dimensional coordinate system.

5. The method of claim 1, wherein said minimum ceiling height is calculated according to the formula:

minimum ceiling height=ceiling offset+A*sin(ceiling adjustment angle),

6. The method of claim 1, further comprising:

a) storing in the memory

(i) a second data structure of two-dimensional coordinates from the floor plan, the two dimensional coordinates in the second data structure representing locations in the physical environment to be avoided by the mobile robot,

(ii) second laser placement information for a second laser attached to the mobile robot, said second laser being oriented to scan the physical environment in a second plane that is not parallel to the floor of the physical environment, wherein said second laser placement information includes a second laser position and a second laser tilt, and

(iii) receiving a second set of laser readings from said second laser, each laser reading corresponding to a location in the physical environment where said second laser detects a physical obstacle in said second plane, and each laser reading comprising a second plane angle and a second plane distance from a given site on the mobile robot;

b) for each laser reading in said second set of laser readings

(i) converting said each laser reading into a second three-dimensional coordinate based on said second plane angle, said second plane distance, said second laser position and said second laser tilt, wherein said second three-dimensional coordinate comprises a second x-component, a second y-component and a second z-component,

(ii) determining a minimum ceiling height and a maximum floor height for the second three-dimensional coordinate based on the second x-component, the second y-component and the tolerance profile, and

(iii) if the second z-component is between the maximum floor height and the minimum ceiling height, adding the second x-component and the second y-component to the second data structure of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot; and

c) causing the propulsion system to drive the mobile robot in the physical environment under control of the first data structure and the second data structure, wherein the propulsion system prevents the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in both the first data structure and the second data structure.

7. The method of claim 6, further comprising:

b) calculating on the mobile robot, in accordance with the floor plan, a path between said current position and said specified location, said path avoiding the locations in the physical environment represented by the first data structure of two-dimensional coordinates and the second data structure of two-dimensional coordinates; and

8. The method of claim 6, wherein said second plane is perpendicular to the floor of the physical environment.

9. In a mobile robot comprising a memory and a propulsion system, a method for avoiding negative obstacles in a physical environment, the method comprising:

a) storing in the memory

(ii) a first data structure of two-dimensional coordinates from the floor plan, the two dimensional coordinates in the first data structure representing locations in the physical environment to be avoided by the mobile robot,

(iv) a tolerance profile comprising a maximum floor offset, a minimum floor offset, a maximum floor adjustment angle, a minimum floor adjustment angle and a maximum allowable floor gap,

b) receiving a set of laser readings from said first laser, each laser reading corresponding to a location in the physical environment within said first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot;

c) for each laser reading in said set of laser readings

(ii) determining a maximum floor height and a minimum floor height for the three-dimensional coordinate based on the x-component, they-component and the tolerance profile, and

(iii) if the z-component is between the maximum floor height and the minimum floor/height, changing an indicator in the memory to designate the x-component and the y-component of the three-dimensional coordinate as the last good floor reading,

(iv) if the z-component is not between the maximum floor height and the minimum floor height, determining whether the distance G between the last good floor reading and the location in the two-dimensional coordinate system defined by the x-component and they-component exceeds the maximum allowable floor gap, and

(v) if the distance G exceeds the maximum allowable floor gap, adding the x-component, the y-component and the last good floor reading to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot;

10. The method of claim 9, further comprising:

11. The method of claim 9, wherein said first plane is perpendicular to the floor of the physical environment.

12. The method of claim 9, wherein said maximum floor height is calculated by using the formula:

maximum floor height=floor offset+A*sin(floor adjustment angle),

13. The method of claim 9, wherein said minimum floor height is calculated according to the formula:

minimum floor height=floor offset−A*sin(floor adjustment angle),

14. A positive obstacle avoidance system for use with a mobile robot in a physical environment, comprising:

a) a first laser attached to the mobile robot, said first laser being oriented to scan the physical environment in a first plane that is not parallel to the floor of the physical environment;

b) a memory for storing

(ii) a first data structure of two-dimensional coordinates from the floor plan, each two dimensional coordinate in the first data structure representing a location in the physical environment to be avoided by the mobile robot,

(iii) first laser placement information for said first laser, said first laser placement information including a first laser position and a first laser tilt, and

c) a laser controller for receiving a set of laser readings from said first laser, each laser reading corresponding to a location in the physical environment where said first laser detects a physical obstacle in said first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot;

d) a positive obstacle avoidance engine that, for each laser reading in said set of laser readings

(i) converts said each laser reading into a three-dimensional coordinate based on said first plane angle, said first plane distance, said first laser position and said first laser tilt, wherein said three-dimensional coordinate comprises an x-component, a y-component and a z-component,

(ii) determines a minimum ceiling height and a maximum floor height for the three dimensional coordinate based on the x-component, they-component and the tolerance profile, and

(iii) if the z-component is between the minimum ceiling height and the maximum floor height, adds the x-component and they-component to the first data structure of two-dimensional coordinates to represent a new location in the physical environment to be avoided by the mobile robot; and

e) a propulsion system that drives the mobile robot in the physical environment under control of the first data structure, the propulsion system being configured to prevent the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in the first data structure.

15. The system of claim 14, further comprising:

a) a communication interface that receives a command to move from a current position to a specified location in the physical environment; and

b) a path planning engine that calculates, in accordance with the floor plan, a path between said current position and said specified location, said path avoiding the locations in the physical environment represented by the first data structure of two-dimensional coordinates;

c) wherein the propulsion system is further configured to automatically drive the mobile robot along the path.

16. The system of claim 14, wherein said first plane is perpendicular to the floor of the physical environment.

17. The system of claim 14, wherein said positive obstacle avoidance engine calculates the maximum floor height according to the formula:

maximum floor height=floor offset+A*sin(floor adjustment angle),

18. The system of claim 14, wherein said positive obstacle avoidance engine calculates the minimum ceiling height according to the formula:

minimum ceiling height=ceiling offset+A*sin(ceiling adjustment angle),

19. The system of claim 14, wherein:

a) a second laser is attached to the mobile robot, said second laser being oriented to scan the physical environment in a second plane that is not parallel to the floor of the physical environment;

b) the memory further comprises

(i) a second data structure of two-dimensional coordinates from the floor plan, each two dimensional coordinate in the second data structure representing another location in the physical environment to be avoided by the mobile robot,

(ii) second laser placement information for said second laser, said second laser placement information including a second laser position and a second laser tilt, and

c) the laser controller receives a second set of laser readings from said second laser, each laser reading corresponding to a location in the physical environment where said second laser detects a physical obstacle in said second plane, and each laser reading comprising a second plane angle and a second plane distance from a given site on the mobile robot;

d) the positive obstacle avoidance engine, for each laser reading in said second set of laser readings

(iii) converts said each laser reading into a second three-dimensional coordinate based on said second plane angle, said second plane distance, said second laser position and said second laser tilt, wherein said second three-dimensional coordinate comprises a second x-component, a second y-component and a second z component,

(iv) determines a minimum ceiling height and a maximum floor height for the second three-dimensional coordinate based on the second x-component, the second y component and the tolerance profile, and

(v) if the second z-component is between the maximum floor height and the minimum ceiling height, adds the second x-component and the second y-component to the second data structure of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot; and

e) the propulsion system drives the mobile robot in the physical environment under control of the first data structure and the second data structure, whereby the propulsion system prevents the mobile robot from passing into the locations in the physical environment represented by the two-dimensional coordinates in both the first data structure and the second data structure.

20. The system of claim 19, further comprising:

21. The system of claim 19, wherein said second plane is perpendicular to the floor of the physical environment.

22. A negative obstacle avoidance system for use with a mobile robot in a physical environment, comprising:

b) a memory for storing

(iii) first laser placement information for said first laser, the first laser placement information including a first laser position and a first laser tilt, and

c) a laser controller for receiving a set of laser readings from said first laser, each laser reading corresponding to a location in the physical environment within said first plane, and each laser reading comprising a first plane angle and a first plane distance from a given site on the mobile robot; and

d) a negative obstacle avoidance engine that, for each laser reading in said set of laser readings

(ii) determines a maximum floor height and a minimum floor height for the three dimensional coordinate based on the x-component, they-component and the tolerance profile, and

(iii) if the z-component is between the maximum floor height and the minimum floor height, changes an indicator in the memory to designate the x-component and the y-component of the three-dimensional coordinate as the last good floor reading coordinate,

(iv) if the z-component is not between the maximum floor height and the minimum floor height, determines whether the distance G between the last good floor reading coordinate and the location in the two-dimensional coordinate system defined by the x-component and they-component exceeds the maximum allowable floor gap, and

(v) if the distance G exceeds the maximum allowable floor gap, adds the x-component, the y-component and the last good floor reading coordinate to the first data structure of two-dimensional coordinates to represent another location in the physical environment to be avoided by the mobile robot; and

23. The system of claim 22, further comprising:

24. The system of claim 22, wherein said first plane is perpendicular to the floor of the physical environment.

25. The system of claim 22, wherein the negative obstacle avoidance engine calculates said maximum floor height by using the formula:

maximum floor height=floor offset+A*sin(floor adjustment angle),

26. The system of claim 22, wherein the negative obstacle avoidance engine calculates said minimum floor height according to the formula:

minimum floor height=floor offset−A*sin(floor adjustment angle),