U.S. patent application number 17/520654 was filed with the patent office on 2022-05-12 for method and apparatus for calculating a dosage of disinfectant applied to an area by an autonomous, mobile robotic device.
The applicant listed for this patent is SAMAN AMARASINGHE, MARCIO MACEDO, ALYSSA PIERSON, DANIELA RUS, YOUSSEF SALEH. Invention is credited to SAMAN AMARASINGHE, MARCIO MACEDO, ALYSSA PIERSON, DANIELA RUS, YOUSSEF SALEH.
Application Number | 20220143250 17/520654 |
Document ID | / |
Family ID | |
Filed Date | 2022-05-12 |
United States Patent
Application |
20220143250 |
Kind Code |
A1 |
PIERSON; ALYSSA ; et
al. |
May 12, 2022 |
METHOD AND APPARATUS FOR CALCULATING A DOSAGE OF DISINFECTANT
APPLIED TO AN AREA BY AN AUTONOMOUS, MOBILE ROBOTIC DEVICE
Abstract
An autonomous, mobile robotic device (AMR) is configured with
one or more UVC radiation sources, and operates to traverse a path
while disinfecting an interior space. Each UVC radiation source is
connected to the AMR by an articulating arm that is controlled to
orient each source towards a feature or surface that is selected
for disinfection during the time that the AMR is moving through the
space. The location of each feature selected for disinfection can
be mapped, and this map information, a current AMR location and
pose can be used to generate signals that are used to control the
articulating arm to orient each UVC lamp towards a feature that is
selected for disinfection.
Inventors: |
PIERSON; ALYSSA;
(SOMERVILLE, MA) ; AMARASINGHE; SAMAN; (WESTON,
MA) ; RUS; DANIELA; (WESTON, MA) ; MACEDO;
MARCIO; (CAMBRIDGE, MA) ; SALEH; YOUSSEF;
(ARLINGTON, MA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
PIERSON; ALYSSA
AMARASINGHE; SAMAN
RUS; DANIELA
MACEDO; MARCIO
SALEH; YOUSSEF |
SOMERVILLE
WESTON
WESTON
CAMBRIDGE
ARLINGTON |
MA
MA
MA
MA
MA |
US
US
US
US
US |
|
|
Appl. No.: |
17/520654 |
Filed: |
November 6, 2021 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
63110784 |
Nov 6, 2020 |
|
|
|
International
Class: |
A61L 2/24 20060101
A61L002/24; G05D 1/02 20060101 G05D001/02; A61L 2/10 20060101
A61L002/10 |
Claims
1. A method of visually representing an amount of UVC radiation
received by an area in an interior space, comprising: determining,
for each of a plurality of points in time, a position and
orientation of an AMR along a path in the interior space; using the
position and orientation of the AMR in the interior space at each
point in time to determining which of a plurality of areas in the
interior space receives UVC radiation emitted by the AMR;
calculating, for each point in time, the amount of UVC radiation
received by each one of the plurality of areas that receives the
UVC radiation; and creating a visual heatmap that represents the
total sum of all UVC radiation received by each of the plurality of
the areas for all of the plurality of points in time.
2. A method of visually modelling a dosage of disinfectant
predicted to be applied to points in an interior space by a UVC
radiation source associated with an autonomous, mobile robotic
device, comprising: generating a map of the interior space using
information collected by a sensor system comprising an autonomous,
mobile, robotic (AMR) device, and determining from the collected
sensor information an identity and location of one or more areas
for disinfection; using the map of the interior space, and the
location of the one or more areas identified for disinfection, to
determine a future path for the AMR device to traverse through the
interior space in order to disinfect each of the one or more areas
identified for disinfection; calculating a dosage of UVC
disinfectant predicted to be applied to a plurality of the points
that are within range of the UVC radiation source as the AMR device
traverses the path; and generating a visual model that illustrates
a degree to which each of the plurality of the points within range
of the UVC radiation source are predicted to receive a dose of
disinfectant by: overlaying on the map the location of each of the
one or more areas identified for disinfection; and plotting on the
map a representation of the degree to which each the plurality of
the points is estimated to receive a dose of disinfectant.
3. The method of claim 2, whereby the dosage of disinfectant is
calculated by considering: a UVC radiation pattern emitted by the
UVC radiation source; a total power of the UVC radiation source; an
orientation of the UVC radiation source with respect to the
interior space; and a distance from the UVC radiation source to
each of the plurality of the points.
4. The method of claim 3, further comprising the dosage of
disinfectant is calculated by considering one or more occlusions
between the UVC radiation source and any of the plurality of the
points comprising the interior space.
5. The method of claim 2, wherein the visual model is a heat map
that illustrates areas of adequate and inadequate doses of
disinfection.
6. The method of claim 5, wherein the areas of adequate and
inadequate doses of disinfection are illustrated by different
colors.
7. The method of claim 5, wherein an area of an adequate dose of
disinfection is illustrated by a color in the red spectrum, and an
area of an inadequate dose of disinfection is illustrated by a
color in the blue spectrum.
8. The method of claim 2, wherein the senor system collects
two-dimensional information or three-dimensional information.
9. The method of claim 2, wherein the areas identified for
disinfection are comprised of either or both of features and
surfaces.
10. The method of claim 2, wherein a calculated value of the
predicted dosage of UVC disinfection is assigned as a weight to an
edge comprising a graph structure used to determine the future path
of the AMR device to through the interior space.
11. The method of claim 10, wherein the weight assigned to the edge
controls a traverse speed of the AMR device and the operation of
the UVC radiation source.
12. A method of visually illustrating a dosage of disinfectant that
is estimated to have been applied to points in an interior space by
a UVC radiation source associated with an autonomous, mobile
robotic device, comprising: generating a map of the interior space
using information collected by a sensor system comprising an
autonomous, mobile, robotic (AMR) device, and determining from the
collected sensor information an identity and location of one or
more areas for disinfection; using the map of the interior space,
and the location of the one or more areas identified for
disinfection, to determine a path for the AMR device to traverse
through the interior space in order to disinfect each of the one or
more areas identified for disinfection; calculating a dosage of
disinfectant estimated to have been applied to a plurality of the
points that are within range of the UVC radiation source as the AMR
device traverses the path; and: generating a visual representation
of the interior space that illustrates a degree to which each of a
plurality of the points within range of the UVC radiation source
are estimated to have receive a dose of disinfectant by: overlaying
on the map the location of each of the one or more areas identified
for disinfection; and plotting on the map a representation of the
degree to which each the plurality of the points is estimated to
have received a dose of disinfectant.
13. The method of claim 12, wherein the dosage of disinfectant is
calculated by considering: a UVC radiation pattern emitted by the
UVC source; a total power of the UVC radiation source; an
orientation of the UVC radiation source with respect to the
interior space; and a distance from the UVC radiation source to
each of the plurality of the points.
14. The method of claim 13, further comprising the dosage of
disinfectant is calculated by considering one or more occlusions
between the UVC radiation source and any of the plurality of the
points comprising the interior space.
15. The method of claim 12, wherein the visual model is a heat map
that illustrates areas of adequate and inadequate doses of
disinfection.
16. The method of claim 15, wherein the areas of adequate and
inadequate doses of disinfection are illustrated by different
colors.
17. The method of claim 15, wherein the areas of adequate doses of
disinfection are illustrated by a color in the red spectrum, and
areas of inadequate doses of disinfection are illustrated by a
color in the blue spectrum.
18. The method of claim 12, wherein the senor system collects
two-dimensional information or three-dimensional information.
19. The method of claim 12, wherein the areas identified for
disinfection are comprised of either or both of features and
surfaces.
20. The method of claim 12, wherein a calculated value of the
predicted dosage of UVC disinfection is assigned as a weight to an
edge comprising a graph structure used to determine the future path
of the AMR device through the interior space.
21. The method of claim 20, wherein the weight assigned to the edge
controls a traverse speed of the AMR device and the operation of
the UVC radiation source.
Description
1. FIELD OF THE INVENTION
[0001] The present disclosure relates to disinfecting an
environment using a UVC radiation source attached to a mobile,
autonomous robotic device.
2. BACKGROUND
[0002] Autonomous, mobile robotic devices (AMRs) can be designed
for use in a variety of different environments for a variety of
different applications. Some AMRs are designed for distribution
center applications, others are designed to clean interior
surfaces, still others are designed to support interaction with
humans in a hospital setting or for audio and/or video
communication applications. Still other AMRs have been designed for
security applications, and some have been designed to rid interior
areas occupied by humans of dangerous pathogens, such as bacteria
and viruses.
[0003] Generally, AMRs designed for the applications mentioned
above can have functionality that enables them to autonomously
follow a predetermined path through their environment, and they can
be designed to have a variety of different types of sensors that
enable them to determine where they are located in their
environment, avoid obstructions in their path, to determine how far
and in what direction they travel, and they can have functionality
that permits them to wireless communicate audio, video or data over
a network with an AMR user.
[0004] With the advent of viral pandemics, the disinfection of
interior spaces has become an important human health issue. In this
regard, there are a number of different disinfection methods being
used to rid spaces of harmful pathogens. For example, disinfecting
liquids or aerosol sprays can be applied to a surface. Other
effective means for disinfecting surfaces are the use of a
filtration or ozonation device, and light sources of particular
wavelengths are also an effective means for disabling many
pathogens. Specifically, electromagnetic radiation having a
wavelength in the 110-280 nm range (i.e., ultraviolet C
radiation/light or UV-C light) is being used to effectively
disinfect surfaces harboring pathogens that are dangerous to human
health. In this regard, hand-held UV-C light sources are available
that allow a user to direct the light source towards an area to be
disinfected while moving around a space. Carts having UV-C light
sources are available that can be manually pushed around to
disinfect a space. More recently, autonomous, mobile robotic
devices having attached UV-C light sources have become available
that can be controlled disinfect interior spaces.
4. BRIEF DESCRIPTION OF THE DRAWINGS
[0005] FIG. 1A is a diagram showing elements comprising a
disinfecting, autonomous, mobile robotic (D-AMR) device 100.
[0006] FIG. 1B is a diagram showing elements comprising a UVC lamp
110a or 110b.
[0007] FIG. 2 is a diagram showing elements comprising a D-AMR
device 200.
[0008] FIG. 3 is a block diagram illustrating functional elements
comprising either of the D-AMR devices 100 or 200.
[0009] FIG. 4 is a block diagram illustrating functional elements
comprising a mapping module 330.
[0010] FIG. 5 is block diagram illustrating functional elements
comprising a path planning and navigation function 340 associated
with either of the D-AMRs 100 and 200.
[0011] FIG. 6 is a block diagram showing a lookup table 345D.
[0012] FIG. 7 is a diagram illustrating a SLAM map 700 with node
locations.
[0013] FIG. 8 is a diagram illustrating the SLAM map of FIG. 7
showing an optimal path from one node to another.
[0014] FIG. 9 is a diagram of a disinfection heat map 900.
5. DETAILED DESCRIPTION
[0015] Generally, autonomous, mobile robotic devices capable of
disinfecting an area in an interior space using UV-C (UVC)
radiation are configured with UVC light or radiation sources which
are fixedly mounted to the AMR in such a manner so as to disinfect
an area within a 360-degree radius around the AMR, and at various
distances (depending upon the power of the light) from the AMR.
This sort of configuration can be dangerous to humans, and other
living things sharing the area being disinfected, as exposure to
UVC radiation for any period of time can be dangerous, and can lead
to skin cancer or eye cataracts. As not all surfaces or features in
an interior space need disinfection, and since a disinfecting
capable AMR, hereinafter referred to as a D-AMR, relies upon
battery power for operation, a D-AMR configuration having
continuous 360-degree disinfection can waste energy and limit
battery life.
[0016] D-AMRs can be provided with a map generated by a user, or
they can automatically generate a map, that identifies
features/surfaces in an area to be disinfected, and the D-AMR can
follow a predetermined path through the mapped space at a constant
rate of speed with the UVC light turned on constantly until it
reaches the end of the path. Provided that a D-AMR typically
traverses a path at a constant rate of speed, and in order to
adequately disinfect all of the surfaces in a given space, it is
necessary to control the rate of speed at which the robot traverses
a path to be relatively slow. This slow rate of speed can be an
unnecessary drain on a battery, and/or require the D-AMR to have a
larger, more expensive battery. Further, a D-AMR user must assume
that a space, after being traversed by the D-AMR, has been
adequately disinfected based upon the rate of speed of the device,
the path of the device, the UVC source intensity and configuration
on the AMR. However, such an assumption that the space has been
adequately disinfected can be erroneous. Further, in the event that
objects in the space have been moved since a disinfection path was
determined, these objects can cast what amounts to a disinfection
shadow, resulting in some surfaces not being disinfected.
[0017] Accordingly, we have designed a D-AMR device configured to
have one or more UVC radiation or light sources attached to it
which operates to disinfect selected features and surfaces in an
interior area, such as a meeting room, cafeteria, or any area that
is commonly occupied by humans or other living things. This D-AMR
traverses a path through an area that is planned based upon a
disinfection need of selected features and surfaces in the area.
These features can be selected by a user of the D-AMR, or they can
be autonomously selected by functionality operating under the
control of the D-AMR.
[0018] According to one embodiment, the speed or speeds at which
the D-AMR moves along at least some portions of a path to disinfect
an interior area can depend upon a configuration of UVC light
sources (i.e., number, type, intensity, and/or orientation of UVC
light sources) attached to the D-AMR. Given a particular UVC light
source configuration, it is possible to calculate a disinfection
dosage model needed to be applied to different features or surfaces
in an area and a disinfection dosage estimate actually applied to
an area in order to substantially neutralize or kill harmful
pathogens living on the features or surfaces. This dosage
information can be utilized by the D-AMR to determine a speed and
whether the UVC lights are turned on or off at different points
along the disinfection path. Further, the path followed by the
D-AMR through the area to be disinfected is determined by the
features to be disinfected, and by the disinfection need (i.e.,
dosage) of each feature. As described herein, a disinfection dosage
can be measured by an amount of UVC radiation applied to and
received by a selected feature over some period of time.
[0019] According to another embodiment, the D-AMR is comprised of a
robot body, drive and sensor systems, and one or more UVC light
sources attached to the robot body. Depending upon the location of
the D-AMR along the disinfection path, the location of features
selected for disinfection, and based upon the desired disinfection
dosage that needs to be applied at particular locations, the
orientation of some or all of UVC light sources comprising a D-AMR
can be controlled independently with respect to an orientation of
the robot body in order to direct UVC radiation towards the
features selected for disinfection. Further, one or all of the UVC
light sources can have reflectors attached to them so that the
radiation emitted by the light sources is directed towards the
features selected for disinfection, and so that the radiation is
directed away from living objects that it can harm (i.e., humans,
plants, and animals).
[0020] According to another embodiment, as the D-AMR traverses a
path through an area, it gathers information that can be used to
generate a 2-D or 3-D heat map that visually illustrates a total
amount of UVC radiation that has been applied over time to
different features comprising the area.
[0021] The aforementioned embodiments will now be described with
reference to the figures, in which FIG. 1A is an illustration of
physical features comprising a D-AMR 100 having two UVC light
sources (hereinafter referred to as UVC lamps) 110a and 110b. Each
of the UVC lamps can have a reflector 130, shown with reference to
FIG. 1B, that functions to direct the UVC radiation towards
features in an interior area that have been selected for
disinfection. The D-AMR 100 has a robot body 115 able to be rotated
around a pivot point 121 with respect to a base 116, and the base
116 has a number of driven wheels 120 that can be controlled by a
drive system to move the D-AMR around an interior area. The D-AMR
can also have a display 1095 and can have one or more sensors 112
that can be employed to, among other things, generate a map of the
space through which the D-AMR is traversing, and to capture images
of objects in the area. Generally, according to one embodiment, the
D-AMR can operate autonomously to generate maps and locate features
in an interior area, and to perform path-planning in the interior
area. The D-AMR can be controlled to move at speeds of between
0.1-1.0 meters per second. As will be described later in more
detail with reference to FIG. 3, a mapping module can have laser
range scanning (Lidar) and simultaneous localization and mapping
functionality (SLAM) that the D-AMR can employ to map and localize
objects in an area and to generate a visual representation of the
map. During operation, the D-AMR can localize its position within a
mapped area using both the lidar functionality and internal
odometry. The D-AMR can also be configured to have one or more
depth cameras used for capturing image information used to identify
different types of objects, and used for dynamic obstacle avoidance
and collision detection. Also, a charging station (not shown) can
be provided that the D-AMR can move to in order to automatically
charge on-board batteries without human intervention. Configured in
such a manner, the D-AMR is able to, independently of human
control, map an area, identify features in the area that can be
disinfected, plan an optimal path through the area and traverse the
path while controlling the orientation of UVC lamps to disinfect
the identified features.
[0022] Further, and as described later, the D-AMR 100 can operate
to create a model to accurately predict a disinfection dosage a
surface in a area will receive from the D-AMR. The dosage
information can be used to control a speed of the D-AMR and to
control the operation of the UVC lamps during the time it traverses
a path through the area. Also, the D-AMR can operate to verify a
dosage received by a surface during the time it traverse a path by
collecting odometry and sensor information. This information
collected by the D-AMR can be used to create a visual heat-map that
provides a user with verification that an area has been
disinfected.
[0023] According to the embodiment shown with reference to FIG. 1A,
the UVC lamps 110a and 110b are shown to be connected to the robot
body by arms 125a and 125b respectively, and each arm can be
controlled to rotate the UVC lamps around an axis that runs along
the length of the arm to cause the UVC radiation to be directed at
different angles with respect to a horizontal plan of a floor
surface 130 in the interior area on which the D-AMR can move.
Rotating the UVC lamps in this manner permits the D-AMR to control
the UVC radiation to be directed to features in the space that are
positioned to be higher or lower with respect to the floor 130. For
example, if a feature to be disinfected is chair having a surface
that is twenty inches above the floor, then the UVC lamps can be
rotated upward or downward as appropriate so that more UVC
radiation is directed towards the chair, or the feature to be
disinfected is a table surface that is thirty inches above the
floor level, then the UVC lamps can be rotated as appropriate so
that more of the UVC radiation is directed towards the table
surface. The ability to control the orientation of the UVC lamps
allows the D-AMR to more effectively and efficiently disinfect
selected features (maybe faster or more efficiently) as it moves
through the interior space. According to another embodiment, the
arms 125A and 125B are not able to be rotated, and have no degrees
of movement freedom. Alternatively, the UVC lamps have reflectors
and can be fixedly mounted to an arm so that rotation with respect
to the arm is not possible. According to this embodiment, the
direction in which the UVC radiation is emitted is controlled by
the orientation of the D-AMR.
[0024] According to another embodiment of the D-AMR 100 in FIG. 1A,
a D-Amr 200 shown with reference to FIG. 2 has UVC lamps 210a and
210b that can be connected to the robot body by respective
articulated robotic arms 225a and 225b, each of which can have
multiple degrees of movement freedom. These articulated arms can be
controlled to orient the lamps so that the UVC radiation is
directed towards selected features in the area being disinfected by
the D-AMR. Further, one or both of the articulated arms can have a
telescopic capability so that the UVC lamps can be extended towards
a feature to be disinfected. Such an articulated arm permits the
D-AMR to more effectively disinfect features in an area.
[0025] Referring now to FIG. 3, which is a diagram illustrating
functional elements comprising either of the D-AMRs 100 or 200, and
which are collectively referred to here as functional modules 300.
A processor 310 is in communication with a non-transitory, computer
memory device (memory) 320 that maintains information generated by
any of the functional modules 300, and which maintains information
that can be used by the processor to control the operation of the
D-AMRs. More specifically, the processor 310 can operate to control
functionality associated with each of the modules 300 to generate
and store information used by the D-AMR to disinfect particular
features in an interior area. The location of the memory 320 with
respect to the D-AMR is not important to the operation of the
D-AMR. The memory device 320 can be on-board the D-AMR as shown in
FIG. 3, or it can be located remotely to the D-AMR in a device
comprising the network 370 or on the user device 380. A mapping
function 330 generally operates to receive information from
different types of sensor devices, and to use this information to
create either 2-D or 3-D maps of the space in which the D-AMR is
operating, and to identify and locate particular features and
surfaces that can be designated for disinfection in an area. A
path-planning and navigation function 340 generally operates on
information generated by the mapping function 330 to determine an
optimal path for the D-AMR to traverse through the area while
performing disinfection, to calculate dosages of UVC radiation to
be applied to features located at different points along the path,
and to determine a particular rate of speed at which the D-AMR can
move along particular portions of the path while disinfecting the
identified features. The module 340 also has functionality that can
operate to control the state of each UVC lamp to be on or off, and
to control the movement of each articulated arm in order to orient
the lamps to be pointing towards a feature identified for
disinfection. The D-AMR also has a wireless communication adapter
350 that operates to receive signals from and send signals to a
wireless access point 360 which operates in conjunction with a
network 370 to route signals to and from the D-AMR. Finally, a
D-AMR user can operate a computational device 380 (i.e., PC or
server) that is connected to the network 360 that can be used to
control or monitor certain aspects of the D-AMR operation.
[0026] Turning now to FIG. 4, which shows the mapping functionality
330 having laser sensing functionality 330A, SLAM functionality
330B, and feature identification functionality 330C. The laser
sensor functionality can be implemented in a 2-D or 3-D light
detection and ranging (Lidar) device, for example, and the output
from this sort of sensor is typically 2-D (x,y) or 3-D (x,y,z)
point cloud information. Information collected from one or more
laser sensing devices 330A can be used by a SLAM (Simultaneous
Location and Mapping) algorithm to construct a visual map showing
the various features comprising an area, such as walls, floors,
tables, chairs, counters, etc. The identify and position of at
least some features comprising the area can be determined by the
feature ID functionality 330C, which can be implemented in a neural
network that is trained to identify features that a user is
interested in disinfecting, such as floors, walls, tables, chairs,
counters, etc. The output of the mapping function 330 can be sent
to the memory 320 where it is stored, and this output is comprised
of a map of the area showing, among other things, the location and
identity of features identified by the feature ID functionality
330C. It should be understood that while the mapping functionality
330 is described as being implemented using a Lidar device, this
mapping function can also be implemented using an image acquired
from cameras or other image sensing devices as opposed to a Lidar
device.
[0027] The path planning and navigation (PPN) module 340 will now
be described with reference to FIG. 5. The PPN module 340 has path
planning functionality 341 and a D-AMR navigation functionality
345. The path planning functionality 341 generally operates under
control of the processor 310 to generate a graph structure by
selecting locations for a plurality of nodes in a mapped area,
identifying all possible paths between the nodes, wherein a path
between any two nodes indicates the existence of an edge,
calculating and assigning a weight to each edge, and determining an
optimum path through the graph. According to one embodiment, a
graph generation function 342 can be controlled to select node
locations independent of any user input, or according to another
embodiment the nodes can be manually selected by a user. Regardless
of whether the position of a node is selected by the function 342
or a user, the position of each node is selected such they can
provide a sufficient dose of disinfection to all features and
surfaces of interest in an area. These nodes are not necessarily
evenly spaced, but can be spaced at a distance such that if a robot
visits a node, it will provide a sufficient disinfection dosage to
the environment surrounding that node. For example, if a room
contains four nodes, by visiting all four nodes, the robot can
provide a sufficient dosage to the entire room. Nodes are also
placed as navigation waypoints between regions of disinfection,
such as navigating down a corridor to an area.
[0028] FIG. 7 shows a SLAM map 700 of an interior area 710 having a
plurality of nodes (illustrated as solid circular objects), each of
which are located at particular selected positions with respect to
the interior area in the map, and two of the nodes are labeled Ni
and Nj. The location of each node comprising the map is fixed by
assigning coordinates to the node. A listing of nodes, and the
connectivity between them (i.e., edge generated by the function
342), defines a graph structure in which each edge can be assigned
a weight that is determined by a dosage of disinfection needed to
be applied to the area between two nodes the edge traverses. For
example, with reference to FIG. 8 which illustrates the same SLAM
map 700 and interior area 710 as shown in FIG. 7, if a dosage of
disinfectant is required at nodes N.sub.i and N.sub.j, a weighting
value of E.sub.ij (the dosage amount) will encode a traversal speed
for the D-AMR from the node Ni to Nj, as well as well as
instructions to control the state of the UVC lights (i.e., one or
off) and, according to one embodiment, information used to control
UVC lamp orientation.
[0029] The PPN module 340 also has graph-search functionality 343
that uses the weighted graph information generated by the graph
generation functionality 342 to find an optimal/shortest path from
one node to another that allows the D-AMR to visit all nodes within
a mapped area and disinfect each feature selected for disinfection.
The graph-search functionality can be implemented in any
appropriate algorithm such as A* or the Traveling Salesman Problem
for example. The information generated by the functions 342 and 343
can be stored in the memory 320 in association with the map
information, and as described above, this information can be
overlayed on the map information to show a correspondence between
features in the map and the weighted graph information.
[0030] Continuing to refer to FIG. 5, the PPN module has
functionality 344 that uses map information generated by the
mapping module 330, path information generated by the path planning
module 341, and UVC light configuration and orientation information
to calculate a disinfection dosage amount that can be applied to
selected features in a mapped area over a period of time that the
D-AMR traverses the path.
[0031] Regarding the dosage calculation function 344, this
functionality is capable of both modeling (344A) and estimating
(344B) the UVC dosages respectively can be applied and that are
applied to surfaces within the environment. The ability to
calculate a dosage applied to each selected feature, allows a
certificate of disinfection to be generated which guarantees that
the specified dosage requirement at each feature or surface in an
area is delivered or not. This disinfection guarantee capability
distinguishes the D-AMR described herein from other mobile UVC
fixtures, which do not have precise localization and mapping, nor
the model of UVC dosage.
[0032] Modeling Dosage Applied by the D-AMR-343A.
[0033] Continuing to refer to FIG. 5, for a particular
configuration of UVC lamps comprising the D-AMR, and using map and
path information, it is possible to calculate a model of UVC
radiation that can be applied to accurately predict a dosage any
selected feature or surface within a particular area can receive. A
simple model of UVC intensity follows an inverse-squared distance
relationship. This relationship is used to estimate the dosage
received from a single UVC lamp within line-of-sight of a point,
and informs the disinfection models based on payload
configurations. The modelling functionality 344A operates to
calculate a predicted disinfection dosage that can be used to
assign a weight to an edge in the weighted graph generated by the
graph-search function 343.
[0034] Dosage Received from a UVC Light Source
[0035] The model of UVC dosage assumes a radially-symmetric light
source, such as a single UVC lamp. Consider a UVC lamp at location
b.sup.i=(x.sub.b, y.sub.b, z.sub.b, .theta..sub.b) in a 3D grid,
where (x.sub.b, y.sub.b, z.sub.b) is the position of the D-AMR and
.theta..sub.b is the heading of the D-AMR. The dosage,
d.sup.i.sub..tau., of UVC light received at a point p=(x, y, z) at
a single time step is expressed in Equation 1 as being proportional
to the inverse squared distance to that point.
d .tau. i .function. ( p , b i ) = .alpha. r 2 .times. .DELTA.
.times. .times. t , Equation .times. .times. 1 ##EQU00001##
where .alpha. is a constant dependent on the power of the light,
r.sup.2=.parallel.b.sup.i-p.parallel..sup.2, and .DELTA.t is an
elapsed time. The units of .alpha. are in Watts, the units of
distance are in centimeters, and the units of dosage
d.sup.i.sub..tau.(p, b.sup.i) are in Watt-seconds per
square-centimeter.
[0036] Dosage Models Based on Robot Configuration
[0037] According to one embodiment, a modular D-AMR payload design
can have multiple UVC lighting configurations that deliver a
disinfection dosage to the environment in a pattern that is more
complex than the
.alpha. r 2 ##EQU00002##
relationship describe above. For a single UVC light source, the
unobstructed dosage will be modified by the payload configuration.
The dosage, D.sup.i.sub..tau.(p, b.sup.i), delivered in Equation 2
below, as
D.sup.i.sub..tau.(p,b.sup.i)=f(d.sup.i.sub..tau.(p,b.sup.i)),
Equation 2
where f (d.sup.i.sub..tau.(p, b.sup.i) is a footprint function.
[0038] The exact form of the footprint function f ( ) will depend
on occlusions between the payload/UVC light configuration, to any
given point. For example, an according to one embodiment, consider
a payload configuration with reflectors, we could choose to write
the disinfection dosage delivered as a piece-wise function,
expressed as in Equation 3 below:
D .tau. i .function. ( p , b i ) = { d .tau. i .function. ( p , b i
) if .times. .times. .phi. 1 .ltoreq. .theta. i .ltoreq. .phi. 2 0
.times. else .times. : Equation .times. .times. 3 ##EQU00003##
[0039] In this example, the disinfection dosage received by a
feature from a UVC lamp light from a payload having a reflector is
defined by where the point (p) associated with the feature is
within the angles (.phi..sub.1, .phi..sub.2) from the robot as
constrained by the reflector and/or the orientation of the UVC
lamp, otherwise, the light will not reach that point.
[0040] To construct a dosage model based on a UVC light
configuration, a summation of all lights on the payload is
considered, their visibility relative to points in the environment
is considered, and the orientation of the robot, or the orientation
of the lamps, within the environment or the orientation of the UVC
lamps is considered. To illustrate how a dosage footprint can be
constructed, let b.sup.i refer to the ith lamp of m total lamps on
a payload. A dosage received at a single timestep can be re-written
as in Equation 4 below:
D.sub..tau.(p,b.sup.1, . . . ,b.sup.m)=.SIGMA..sub.i=1.sup.m
D.sup.i.sub..tau.(p,b.sup.i), Equation 4
where the total dosage delivered from the payload D.sub..tau. is
the summation of the dosage delivered by each individual light
D.sup.i.sub..tau..
[0041] Computing Dosage, D(p, b), Accumulation for a Mobile Robot
Over Time
[0042] If the D-AMR is stationary, then the dose can be calculated
in one step by the total elapsed time. However, as the D-AMR moves,
the distance between the D-AMR and the point to be disinfected
changes. The computation of the total accumulated dosage at a point
as the sum of all dosages at each time step is expressed in
Equation 5 below:
D(p,b)=.SIGMA..sub..tau.=1.sup.n D.sub..tau.(p.sub.t,b.sub.t),
Equation 5
wherein, .tau. represents a discretized time step, and n is the
total number of steps. If our recorded robot trajectory has 100
points, then n=100, and for each unique point in the trajectory, a
D.sub..tau. calculation is performed using Equation 4. Note that
.DELTA.t can be computed as the elapsed time between trajectory
samples. The expression D(p, b) provides an estimate of the
accumulated dosage at a point in the environment over time for a
mobile robot moving through the space. Here, D.sub..tau.(p.sub.t,
b.sub.t) represents the dosage pattern from the robot for a single
time step. The design of D.sub..tau.(p.sub.t, b.sub.t) should
follow from the configuration of the UVC lights within the payload,
and different payload configurations will have different dosage
models.
[0043] Verifying Dosage Received in 2D and 3D
[0044] Referring now to the estimating function 344B in FIG. 5,
during each traverse of a path, the D-AMR collects odometry
information, sensor data, and applied dosage information generated
using Equation 5. From the trajectory history and map information,
the function 344B can estimate a total dosage of UVC radiation that
was applied to all surfaces within the environment. For each point
in time, there is a corresponding position of the D-AMR within a
map, as well as any sensor information (such as camera and lidar
information). To accurately calculate and verify the dosage, the
D-AMR relies on sensor data to account for occlusions within the
area along that path. For instance, since lidar gives distances to
walls and objects in the environment, it is a natural proxy for the
line-of-sight coverage of the UVC dosages. It is possible to
reconstruct a footprint of dosage from the lidar information at
each point in time along the path to the points known to be visible
to the D-AMR at that point in time. With a 2D lidar device, it is
possible to create this map flattened in the z-dimension.
Information generated by a 3D lidar device can be used to
reconstruct a 3D walk-through of the environment, with the known
dosage at every point in the 3D space.
[0045] Visualizing Dosage Applied in 2D and 3D
[0046] Subsequent to dosages delivered in an area being verified by
the estimating function 344B, a map visualizing the effectiveness
of the disinfection process can be generated. Using information
generated by the estimating function 344B, a 2D heatmap can be
generated having a color gradient that denotes the various dosage
thresholds delivered by the robot in time. Color can range from red
to blue, with red representing an area in which an adequate or high
dosage of disinfectant is applied and blue representing an area in
which a low or inadequate dosage of disinfectant is applied. This
heatmap can be overlaid onto a building floorplan or a SLAM map to
create a dosage report for the end user, where all points in the
robot's SLAM map are translated into the coordinates of the end
user's map. Such a heatmap 900 is illustrated with reference to
FIG. 9. More specifically, since the disinfection dosage applied to
each location (i.e., location defined by 2-D or 3-D coordinates
(x,y) or (x,y,z) respectively) in a space is known, these locations
can be plotted and then overlayed onto a map of the space in a
manner that shows a gradient change from one area of greater
disinfection to an area of lesser disinfection. This heatmap can be
implemented in python or any other appropriate scripting
language.
[0047] In 3D, lidar or camera information recorded by the D-AMR
during a path traverse is sufficient to generate a 3D
reconstruction of an area. Within the 3D reconstruction, the
surfaces of this model can be colored, with the colormap
corresponding to the overall dosage received. An end user would be
able to either explore the model in an interactive fashion, or a
simulation replaying the D-AMRs trajectory in the modeled
environment can be provided to the user.
[0048] Referring again to FIG. 5, the D-AMR navigation
functionality 345 generally operates to examine information
comprising a map to generate navigation instructions that the D-AMR
can use to move from one node, or waypoint, to the next node along
a path through an area to be disinfected. The navigation function
345 has a path tracking function 345A, pose determination function
345B and D-AMR arm articulation control functionality 345C. The
path tracking function 345A can be implemented in any appropriate
algorithm that operates to calculate an angular velocity from a
current, known location (i.e., a current node location), to a next
node location along a path using the information generated by the
path planning function 34. In order to operate, the tracking
function uses information generated by the pose function 345B,
which in turn operates using a current location of the D-AMR (which
can be determined using odometer information) and sensor
information, such a lidar or other sensor information, to determine
the current location and pose (orientation within a mapped space).
This pose information can be maintained until used by the tracking
function. The odometer information can be received from an odometer
device operating in association with a D-AMR drive system (not
shown), and this information is generally saved as a distance
travelled in feet, for example, from a starting location to the
current location of the D-AMR.
[0049] Continuing to refer to FIG. 5, the speed at which the D-AMR
moves from one node to another can be determined by an amount of
disinfection that is needed as the D-AMR traverses a path between
the nodes. And this traversal speed corresponds to a disinfection
dosage assigned to an edge, as a weight, that corresponds to the
path between the two nodes. This weight encodes a particular speed
at which the D-AMR moves along the path between the two nodes and
it encodes the state of the UVC lamps comprising the D-AMR. The
encoded information associated with the speed and lamp control can
be contained in a table 345D shown with reference to FIG. 6. The
speed and lamp state information comprising the Table 345D is
generated by the path planning module 340. As shown in in the
table, each of three doses, A, B and C, calculated by the path
planning module 340 is associated with a speed (in meters per
second) and a lamp state (on or off). Alternatively, the speeds and
lamp state information can be generated manually by an expert user
based on a dosage requirement and a type of feature to be
disinfected in an area. And, according to one embodiment,
instructions used to control the orientation of a UVC lamp
comprising the D-AMR can also be associated with each edge. In this
regard, knowing a starting orientation for each D-AMR arm, the
orientation of the D-AMR in the mapped area (from the pose
information), and knowing the location of a feature selected for
disinfection with respect to the current location of the D-AMR, the
functionality 345C can operate to generate arm articulation
instructions that can be used to control the arm movement to direct
a UVC lamp towards a feature in the area selected for disinfection.
An arm can operate to raise or lower a lamp with respect to a floor
plane, it can operate to rotate a lamp around a horizontal or
vertical axis alter the direction of the UVC radiation emitted from
the lamp, and it can operate to extend a lamp away from the D-AMR
body towards a feature to be disinfected.
[0050] Regarding the functionality 341 that operates to select that
location or locations and identities of each of one or more nodes
comprising the map, this functionality can be controlled by the
D-AMR to run independently to select node locations, or the node
locations can be selected under user control. Regardless of whether
the position of nodes is selected by the D-AMR or user, the
position of each node is selected such they can provide a
sufficient dose of disinfection to all features, and surfaces of
interest comprising an area. These nodes are not necessarily evenly
spaced, but instead spaced at a distance such that if a robot
visits a node, it will provide sufficient dosage to the environment
surrounding that node. For example, if a room contains four nodes,
by visiting all four nodes, the robot provides sufficient dosage to
the entire room. Nodes are also placed as navigation waypoints
between regions of disinfection, such as navigating down a corridor
to an area.
* * * * *