Method and system of sensing the best-connected future path for a mobile telerobot
This disclosure relates generally to a method and system for sensing best-connected future path for a mobile telerobot based on radio signal strength (RSS) prediction algorithm through in-situ radio-sensing. State-of-the-art methods predict the future path from the plurality of possible paths based on a radio-source in the environment. However, prediction of the suitable future path in the absence of the radio-source or in no signal zone is not yet achieved. The proposed in-situ algorithm is based on Log-Normal Shadowing Model (LNSM) and found efficient for prediction error minimization. The method enables the mobile telerobot to predict the future path on a trajectory of the telerobot even without prior knowledge of a radio-source location. The mobile telerobot can predict the most suitable path from a plurality of possible paths for a move based on virtual location estimation.
1 . A method of sensing best-connected future path for a mobile telerobot, the method comprising:
receiving, via one or more hardware processors, an odometry and a corresponding received radio signal strength (RSS) measurement at a current location of the mobile telerobot, wherein the odometry comprises information data used by a motion sensor of the mobile telerobot to estimate change in position over time, wherein the mobile telerobot is controlled by a remote operator via a controller communicating with the mobile telerobot, wherein the mobile telerobot take a plurality of possible paths to reach a destination point from a source location;
estimating, via the one or more hardware processors, a virtual access point (AP) location from the current location of the mobile telerobot comprising steps:
obtaining a first virtual AP location followed by estimating subsequent virtual AP locations, wherein the subsequent virtual AP locations are estimated until error between predicted RSS value and received RSS value at the current point remains within a pre-defined threshold in a trajectory of the mobile telerobot by,
revolving the mobile telerobot 360 degrees at the current position to obtain raw RSS values by capturing the RSS across 360 degrees that are influenced by the environmental factors including static and dynamic obstacles along telerobot's trajectory,
filtering the raw RSS values to obtain filtered RSS values through Butterworth order low pass filter with predefined cutoff frequency, determining a burst maximum of the filtered RSS values and corresponding peak to obtain a peak angle, and
estimating the first virtual AP location using the peak angle;
determining, via the one or more hardware processors, a subsequent virtual AP location using a Trilateration algorithm, for a prediction error at the current point exceeds a pre-defined threshold, wherein the Trilateration algorithm uses three reference nodes to calculate the positions of unknown node, and uses a Received Signal Strength Indicator (RSSI) value at points previously traversed by the mobile robot along its trajectory to estimate the subsequent virtual AP location, wherein the first AP location becomes an input parameter to the Trilateration algorithm, wherein the prediction error is computed for each advancing move based on the RSSI obtained at the point where the mobile telerobot moves and the RSSI value for that current point is predicted at that point using the current RSS prediction model;
iteratively obtaining, via the one or more hardware processors, subsequent virtual AP locations as the mobile telerobot advances to the next move upon its traversal, wherein the subsequent virtual AP location is calculated at the current point for the prediction at that point exceeds the pre-defined threshold based on the estimated location of virtual AP;
calculating, via the one or more hardware processors, a virtual distance between the virtual AP location and a future point wherein the future point is the set of coordinates of an unvisited point along the trajectory of the mobile telerobot P steps ahead from the current positions of the mobile telerobot, wherein the unvisited future point on the trajectory of the mobile telerobot is identified based on (i) a current pose of the telerobot, (ii) a direction of the future point and (iii) the co-ordinates of the current point of the mobile telerobot;
wherein the virtual AP location is calculated using an equation:
xa
=
d
sin
(
α
)
ya
=
d
cos
(
α
)
wherein d is an Euclidean distance between the first location and the AP determined using the received RSS and wherein the Euclidean distance is calculated by equation:
d
=
antilog
10
RSS
d
0
-
R
S
S
d
,
t
10
η
where, d0 is the reference distance usually 1 m, RSSd0 is the received RSS at a reference distance, η is an environment dependent path loss parameter, and α is a peak angle corresponding to burst maximum values of filtered RSS obtained in the previous steps;
feeding, via the one or more hardware processors, the virtual distance into a path loss model to predict the RSS at the future point, wherein the predicted RSS at the future point assists the remote operator navigating the mobile telerobot via the controller to choose best-connected path from the plurality of possible paths; and
identifying, via one or more hardware processors, best-connected path from the plurality of possible paths based on the predicted RSS at the future point, wherein at any point in telerobot's trajectory, prediction of RSS at future, unvisited points along the possible paths to destination helps the remote operator to decide which path to take and the telerobot standing at a point makes P steps ahead prediction along different directions based on the calculated RSS at the future point which assists the remote operator to choose the best-connected path among the plurality of possible paths for the mobile telerobot.
2 . The method as claimed in claim 1 , wherein the first estimated virtual AP location, an odometry and the RSS of at least three points previously traversed by the mobile telerobot are provided as input parameters to a first round of Trilateration algorithm, wherein previously estimated virtual AP location obtained from the previous iteration of Trilateration algorithm, an odometry and the RSS information of at least three points previously traversed by the mobile telerobot are provided as input parameters to a subsequent iterations of the Trilateration algorithm.
3 . The method as claimed in claim 1 , wherein filtering of raw RSS involves passing the raw RSS through Butterworth order low pass filter with pre-defined cutoff frequency.
4 . The method as claimed in claim 1 wherein the Trilateration algorithm is applied to predict the virtual AP location using the following set of equations:
d
1
=
(
x
a
-
x
1
)
2
+
(
y
a
-
y
1
)
2
d
2
=
(
x
a
-
x
2
)
2
+
(
y
a
-
y
2
)
2
d
3
=
(
x
a
-
x
3
)
2
+
(
y
a
-
y
3
)
2
wherein (xa,ya) is the estimated location of virtual AP, (xi,yi) corresponds of each point i already traversed by the mobile telerobot along its trajectory and d i is the corresponding Euclidean distance of point i from the radio source computed using the RSS received at each point i.
5 . The method as claimed in claim 1 , wherein the coordinates of the future point are calculated using equation:
x
p
=
x
+
(
P
*
sin
(
θ
+
β
)
)
y
p
=
y
+
(
P
*
cos
(
θ
+
β
)
)
wherein the future point is assumed to be P steps ahead from the current location (x,y), and in β direction with respect to the current pose of the mobile telerobot θ.
6 . A system, comprising:
a memory storing instructions;
one or more communication interfaces; and
one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to:
receive an odometry and a corresponding RSS measurement at a current location of the mobile telerobot, wherein the odometry comprises information data used by a motion sensor of the mobile telerobot to estimate change in position over time, wherein the mobile telerobot is controlled by a remote operator via a controller communicating with the mobile telerobot, wherein the mobile telerobot take a plurality of possible paths to reach a destination point from a source location;
estimate a virtual access point (AP) location from the current location of the mobile telerobot comprising:
obtaining a first virtual AP location followed by estimating subsequent virtual AP locations, wherein the subsequent virtual AP locations are estimated until error between predicted RSS value and received RSS value at the current point remains within a pre-defined threshold in a trajectory of the mobile telerobot by,
revolving the mobile telerobot 360 degrees at a current position to obtain the raw received radio signal strength (RSS) values by capturing the RSS across 360 degrees that are influenced by the environmental factors including static and dynamic obstacles along telerobot's trajectory, filtering the raw RSS values to obtain filtered RSS values through Butterworth order low pass filter with predefined cutoff frequency,
determining a burst maximum of the filtered RSS values and corresponding peak to obtain a peak angle,
estimating the first virtual AP location using the peak angle;
determine a subsequent virtual AP location using a Trilateration algorithm, for a prediction error at the current point exceeds a pre-defined threshold, wherein the Trilateration algorithm uses three reference nodes to calculate the positions of unknown node, and uses a Received Signal Strength Indicator (RSSI) value at points previously traversed by the mobile robot along its trajectory to estimate the subsequent virtual AP location, wherein the first AP location becomes an input parameter to the Trilateration algorithm, wherein the prediction error is computed for each advancing move based on the RSSI obtained at the point where the mobile telerobot moves and the RSSI value for that current point is predicted at that point using the current RSS prediction model;
iteratively obtain subsequent virtual AP locations as the mobile telerobot advances to the next move upon its traversal, wherein the subsequent virtual AP location is calculated at the current point if the prediction at that point exceeds the pre-defined threshold;
calculate a virtual distance between the virtual AP location and a future point wherein the future point is the set of coordinates of an unvisited point along the trajectory of the mobile telerobot P steps ahead from the current positions of the mobile telerobot, wherein the unvisited future point on the trajectory of the mobile telerobot is identified based on (i) a current pose of the telerobot, (ii) a direction of the future point and (iii) the co-ordinates of the current point of the mobile telerobot;
wherein the virtual AP location is calculated using an equation:
xa
=
d
sin
(
α
)
ya
=
d
cos
(
α
)
wherein d is an Euclidean distance between the first location and the AP determined using the received RSS and wherein the Euclidean distance is calculated by equation:
d
=
antilog
10
R
S
S
d
0
-
R
S
S
d
,
t
10
η
where, d0 is the reference distance usually 1 m, RSSd0 is the received RSS at a reference distance, η is an environment dependent path loss parameter, and α is a peak angle corresponding to burst maximum values of filtered RSS obtained in the previous steps;
feed the virtual distance into a path loss model to predict the RSS at the future point;
wherein the predicted RSS at the future point assists a remote operator navigating the mobile telerobot to choose the best-connected path from the plurality of possible paths; and
identify, best-connected path from the plurality of possible paths based on the predicted RSS at the future point, wherein at any point in telerobot's trajectory, prediction of RSS at future, unvisited points along the possible paths to destination helps the remote operator to decide which path to take and the telerobot standing at a point makes P steps ahead prediction along different directions based on the calculated RSS at the future point which assists the remote operator to choose the best-connected path among the plurality of possible paths for the mobile telerobot.
7 . The system of claim 6 , wherein the first estimated virtual AP location, an odometry and the RSS of at least three points previously traversed by the mobile telerobot are provided as input parameters to a first round of Trilateration algorithm, wherein previously estimated virtual AP location obtained from the previous iteration of Trilateration algorithm, an odometry and the RSS information of at least three points previously traversed by the mobile telerobot are provided as input parameter to a subsequent iterations of the Trilateration algorithm.
8 . The system of claim 6 , wherein filtering of raw RSS involves passing the raw RSS through Butterworth order low pass filter with pre-defined cutoff frequency.
9 . The system of claim 6 , wherein the Trilateration algorithm is applied to predict the virtual AP location using the following set of equations:
d
1
=
(
x
a
-
x
1
)
2
+
(
y
a
-
y
1
)
2
d
2
=
(
x
a
-
x
2
)
2
+
(
y
a
-
y
2
)
2
d
3
=
(
x
a
-
x
3
)
2
+
(
y
a
-
y
3
)
2
wherein (xa,ya) is the estimated location of virtual AP, (x i ,y i ) corresponds of each point i already traversed by the mobile telerobot along its trajectory and d i is the corresponding Euclidean distance of point i from the radio source computed using the RSS received at each point i.
10 . The system of claim 6 , wherein the coordinates of the future point are calculated using equation:
x
p
=
x
+
(
P
*
sin
(
θ
+
β
)
)
y
p
=
y
+
(
P
*
cos
(
θ
+
β
)
)
wherein the future point is assumed to be P steps ahead from the current location (x,y), and in β direction with respect to the current pose of the mobile telerobot θ.
11 . One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:
receiving an odometry and a corresponding received radio signal strength (RSS) measurement at a current location of the mobile telerobot, wherein the odometry comprises information data used by a motion sensor of the mobile telerobot to estimate change in position over time, wherein the mobile telerobot is controlled by a remote operator via a controller communicating with the mobile telerobot, wherein the mobile telerobot take a plurality of possible paths to reach a destination point from a source location;
estimating a virtual access point (AP) location from the current location of the mobile telerobot comprising steps:
obtaining a first virtual AP location followed by estimating subsequent virtual AP locations, wherein the subsequent virtual AP locations are estimated until error between predicted RSS value and received RSS value at the current point remains within a pre-defined threshold in a trajectory of the mobile telerobot by,
revolving the mobile telerobot 360 degrees at the current position to obtain raw RSS values by capturing the RSS across 360 degrees that are influenced by the environmental factors including static and dynamic obstacles along telerobot's trajectory,
filtering the raw RSS values to obtain filtered RSS values through Butterworth order low pass filter with predefined cutoff frequency, and
determining a burst maximum of the filtered RSS values and corresponding peak to obtain a peak angle,
estimating the first virtual AP location using the peak angle;
determining a subsequent virtual AP location using a Trilateration algorithm, for a prediction error at the current point exceeds a pre-defined threshold, wherein the Trilateration algorithm uses three reference nodes to calculate the positions of unknown node, and uses a Received Signal Strength Indicator (RSSI) value at points previously traversed by the mobile robot along its trajectory to estimate the subsequent virtual AP location, wherein the first AP location becomes an input parameter to the Trilateration algorithm, wherein the prediction error is computed for each advancing move based on the RSSI obtained at the point where the mobile telerobot moves and the RSSI value for that current point is predicted at that point using the current RSS prediction model;
iteratively obtaining subsequent virtual AP locations as the mobile telerobot advances to the next move upon its traversal, wherein the subsequent virtual AP location is calculated at the current point for the prediction at that point exceeds the pre-defined threshold based on the estimated location of virtual AP;
calculating a virtual distance between the virtual AP location and a future point wherein the future point is the set of coordinates of an unvisited point along the trajectory of the mobile telerobot P steps ahead from the current positions of the mobile telerobot, wherein the unvisited future point on the trajectory of the mobile telerobot is identified based on (i) a current pose of the telerobot, (ii) a direction of the future point and (iii) the co-ordinates of the current point of the mobile telerobot;
wherein the virtual AP location is calculated using an equation:
xa
=
d
s
in
(
α
)
ya
=
d
cos
(
α
)
wherein d is an Euclidean distance between the first location and the AP determined using the received RSS and wherein the Euclidean distance is calculated by equation:
d
=
antilog
10
R
S
S
d
0
-
R
S
S
d
,
t
10
η
where, d0 is the reference distance usually 1 m, RSSd0 is the received RSS at a reference distance, η is an environment dependent path loss parameter, and α is a peak angle corresponding to burst maximum values of filtered RSS obtained in the previous steps;
feeding the virtual distance into a path loss model to predict the RSS at the future point, wherein the predicted RSS at the future point assists a remote operator navigating the mobile telerobot via a controller to choose the best-connected path from the plurality of possible paths; and
identifying, via one or more hardware processors, best-connected path from the plurality of possible paths based on the predicted RSS at the future point, wherein at any point in telerobot's trajectory, prediction of RSS at future, unvisited points along the possible paths to destination helps the remote operator to decide which path to take and the telerobot standing at a point makes P steps ahead prediction along different directions based on the calculated RSS at the future point which assists the remote operator to choose the best-connected path among the plurality of possible paths for the mobile telerobot.
12 . The one or more non-transitory machine-readable information storage mediums of claim 11 , wherein the first estimated virtual AP location, an odometry and the RSS of at least three points previously traversed by the mobile telerobot are provided as input parameters to a first round of Trilateration algorithm, wherein previously estimated virtual AP location obtained from the previous iteration of Trilateration algorithm, an odometry and the RSS information of at least three points previously traversed by the mobile telerobot are provided as input parameter to a subsequent iterations of the Trilateration algorithm.
13 . The one or more non-transitory machine-readable information storage mediums of claim 11 , wherein filtering of raw RSS involves passing the raw RSS through Butterworth order low pass filter with pre-defined cutoff frequency.
14 . The one or more non-transitory machine-readable information storage mediums of claim 11 , wherein the Trilateration algorithm is applied to predict the virtual AP location using the following set of equations:
d
1
=
(
x
a
-
x
1
)
2
+
(
y
a
-
y
1
)
2
d
2
=
(
x
a
-
x
2
)
2
+
(
y
a
-
y
2
)
2
d
3
=
(
x
a
-
x
3
)
2
+
(
y
a
-
y
3
)
2
wherein (xa,ya) is the estimated location of virtual AP, (xi,yi) corresponds of each point i already traversed by the mobile telerobot along its trajectory and d i is the corresponding Euclidean distance of point i from the radio source computed using the RSS received at each point i.
15 . The one or more non-transitory machine-readable information storage mediums of claim 11 , wherein the coordinates of the future point are calculated using equation:
x
p
=
x
+
(
P
*
sin
(
θ
+
β
)
)
y
p
=
y
+
(
P
*
cos
(
θ
+
β
)
)
wherein the future point is assumed to be P steps ahead from the current location (x,y), and in β direction with respect to the current pose of the mobile telerobot θ.