Dataset simplification of multidimensional signals captured for asset tracking
Methods, systems, and devices for dataset simplification of multidimensional signals captured for asset tracking are provided. An example method involves obtaining raw data from a data source onboard an asset and determining whether obtainment of the raw data results in satisfaction of a data logging trigger. The method further involves, when the data logging trigger is satisfied, performing a dataset simplification algorithm on a target set of data within the raw data to generate a simplified set of data, wherein the target set of data contains a time-variant multidimensional signal and the dataset simplification algorithm is generalized for any multidimensional signal. The method further involves transmitting the simplified set of data to a server.
1. A method comprising:
obtaining, at a telematics device onboard an asset, the asset being a vehicle or associated with a vehicle, raw data;
when a data logging trigger is satisfied, with the telematics device onboard the asset, reducing a target set of data within the raw data into a simplified set of data, wherein:
the target set of data contains at least a first dimension for a sequence of time values, a second dimension for a first type of data collected with respect to the sequence of time values, and a third dimension for a second type of data collected with respect to the sequence of time values; and
reducing the target set of data into the simplified set of data comprises applying, to the target set of data, a modified Ramer-Douglas-Peucker algorithm in which points are selected from the target set of data for inclusion in the simplified set of data based on distances of the points to iteratively-defined reference lines defined through portions of the target set of data in n-dimensional space, where n is three or greater; and
transmitting the simplified set of data from the telematics device to a server for provision of a telematics service.
2. The method of claim 1 , wherein:
the distances of the points to the iteratively-defined reference lines are minimum distances in the n-dimensional space where n is three or greater; and
determination of a minimum distance for a sample point in the target set of data to a reference line comprises:
determining a reference point on the reference line that is minimally distant from that sample point by evaluating t min in the following equation:
t
p
+
∑
n
=
1
N
(
y
p
n
-
y
0
n
)
*
a
n
1
+
∑
n
=
1
N
a
n
2
wherein,
t p represents the time value (t) at the sample point (p),
N represents the total number of non-time dimensions in the multidimensional signal,
n represents the n th non-time dimension in the multidimensional signal,
y p n represents the value (y) of the sample point (p) in the n th dimension (n),
y 0 n represents the value (y) of the first point (0) in the n th dimension (n), the first point (0) being first with respect to time,
a n represents the slope (a) of the reference line in the n th dimension (n) with respect to time, and
t min represents the time (t) at the reference point, the point at which the reference line is minimally distant (min) from the sample point; and
determining the distance from that sample point to the reference point by the following equation:
R
(
L
,
p
)
=
(
t
min
-
t
p
)
2
+
∑
n
=
1
N
(
y
0
n
+
a
n
*
t
min
-
y
p
n
)
2
wherein,
t min , t p , N, n, y 0 n , a n , and y p n are defined as in the previous equation, and
|R(L,p)| represents the distance (R) from the sample point (p) to the reference point on the reference line (L).
3. The method of claim 1 , wherein:
the first type of data and the second type of data are recorded in different units, and
the method further comprises scaling the units of one or both of the first type of data and the second type of data prior to reduction of the target set of data into the simplified set of data.
4. The method of claim 1 , wherein the first type of data includes a latitudinal position of the asset and the second type of data includes a longitudinal position of the asset.
5. The method of claim 4 , wherein the target set of data further includes a fourth dimension for a third type of data that includes an altitudinal position of the asset.
6. The method of claim 1 , wherein:
the first type of data includes an accelerometer signal in an X direction,
the second type of data includes an accelerometer signal in a Y direction,
the target set of data further includes a fourth dimension for a third type of data, and
the third type of data includes an accelerometer signal in a Z direction.
7. The method of claim 1 , wherein the raw data is obtained from a data source that includes one or more of: an onboard diagnostic port of the asset or a sensor of the asset tracking device.
8. The method of claim 1 , wherein the telematics device is integrated into the asset.
9. A telematics device for use onboard an asset, the telematics device comprising:
an interface layer to obtain raw data from a data source onboard the asset, the asset being a vehicle or associated with a vehicle;
a memory to store the raw data;
a controller to:
when a data logging trigger is satisfied with the controller of the telematics device, reduce a target set of data within the raw data into a simplified set of data, wherein:
the target set of data contains at least a first dimension for a sequence of time values, a second dimension for a first type of data collected with respect to the sequence of time values, and a third dimension for a second type of data collected with respect to the sequence of time values; and
reducing the target set of data into the simplified set of data involves applying, to the target set of data a modified Ramer-Douglas-Peucker algorithm in which points are selected from the target set of data for inclusion in the simplified set of data based on distances of the points to iteratively-defined reference lines defined through portions of the target set of data in n-dimensional space, where n is three or greater; and
a communication interface to transmit the simplified set of data from the telematics device to a server for provision of a telematics service.
10. The telematics device of claim 9 , wherein the first type of data and the second type of data are recorded in different units, and the units of one or both of the first type of data and the second type of data are scaled prior to reduction of the target set of data into the simplified set of data.
11. The telematics device of claim 9 , wherein:
the distances are minimum distances in n-dimensional space, n being three or greater; and
determination of a minimum distance for a sample point in the target set of data to a reference line comprises:
determining a reference point on the reference line that is minimally distant from that sample point by evaluating t min in the following equation:
t
min
=
t
p
+
∑
n
=
1
N
(
y
p
n
-
y
0
n
)
*
a
n
1
+
∑
n
=
1
N
a
n
2
wherein,
t p represents the time value (t) at the sample point (p),
N represents the total number of non-time dimensions in the target set of data, where N is two or greater,
n represents the n th non-time dimension in the target set of data,
y p n represents the value (y) of the sample point (p) in the n th dimension (n),
y 0 n represents the value (y) of the first point (0) in the n th dimension (n), the first point (0) being first with respect to time,
a n represents the slope (a) of the reference line in the n′h dimension (n) with respect to time, and
t min represents the time (t) at the reference point, the point at which the reference line is minimally distant (min) from the sample point; and
determining the distance from that sample point to the reference point by the following equation:
R
(
L
,
p
)
=
(
t
min
-
t
p
)
2
+
∑
n
=
1
N
(
y
0
n
+
a
n
*
t
min
-
y
p
n
)
2
wherein,
t min , t p , N, n, y 0 n , a n , and y p n are defined as in the previous equation, and
|R(L,p)| represents the distance (R) from the sample point (p) to the reference point on the reference line (L).
12. The telematics device of claim 9 , wherein the telematics device is integrated into the asset.
13. The telematics device of claim 9 , wherein the first type of data includes a latitudinal position of the asset and the second type of data includes a longitudinal position of the asset.
14. The telematics device of claim 9 , wherein:
the first type of data includes an accelerometer signal in an X direction,
the second type of data includes an accelerometer signal in a Y direction,
the target set of data further includes a fourth dimension for a third type of data; and
the third type of data includes an accelerometer signal in a Z direction.
15. A system comprising:
a telematics device onboard an asset, the asset being a vehicle or associated with a vehicle, the telematics device configured to:
obtain raw data from a data source onboard the asset;
when a data logging trigger is satisfied, with the telematics device, reduce a target set of data within the raw data into a simplified set of data, wherein:
the target set of data contains at least a first dimension for a sequence of time values, a second dimension for a first type of data collected with respect to the sequence of time values, and a third dimension for a second type of data collected with respect to the sequence of time value;
reducing the target set of data into the simplified set of data involves applying, to the target set of data, a modified Ramer-Douglas-Peucker algorithm in which points are selected from the target set of data for inclusion in the simplified set of data based on distances of the points to iteratively-defined reference lines defined through portions of the target set of data in n-dimensional space, n being three or greater; and
transmit the simplified set of data from the telematics device;
one or more servers to:
receive the simplified set of data from the telematics device; and
record the simplified set of data in an asset tracking database for provision of a telematics service.
16. The system of claim 15 , wherein the first type of data and the second type of data are recorded in different units, and one or both of the first type of data and the second type of data are scaled prior to reduction of the target set of data into the simplified set of data.
17. The system of claim 15 , wherein the telematics device is integrated into the asset.
18. The system of claim 15 , wherein:
the distances are minimum distances in the n-dimensional space, n being three or greater, and
determination of a minimum distance for a sample point in the target set of data to a reference line comprises:
determining a reference point on the reference line that is minimally distant from that sample point by evaluating t min in the following equation:
t
min
=
t
p
+
∑
n
=
1
N
(
y
p
n
-
y
0
n
)
*
a
n
1
+
∑
n
=
1
N
a
n
2
wherein,
t p represents the time value (t) at the sample point (p),
N represents the total number of non-time dimensions in the target set of data, where N is two or greater,
n represents the n th non-time dimension in the target set of data,
y p n represents the value (y) of the sample point (p) in the n th dimension (n),
y 0 n represents the value (y) of the first point (0) in the n th dimension (n), the first point (0) being first with respect to time,
a n represents the slope (a) of the reference line in the nt h dimension (n) with respect to time, and
t min represents the time (t) at the reference point, the point at which the reference line is minimally distant (min) from the sample point; and
determining the distance from that sample point to the reference point by the following equation:
R
(
L
,
p
)
=
(
t
min
-
t
p
)
2
+
∑
n
=
1
N
(
y
0
n
+
a
n
*
t
min
-
y
p
n
)
2
wherein,
t min , t p , N, n, y 0 n , a n , and y p n are defined as in the previous equation, and
|R(L,p)| represents the distance (R) from the sample point (p) to the reference point on the reference line (L).
19. The system of claim 15 , wherein the first type of data includes a latitudinal position of the asset and the second type of data includes a longitudinal position of the asset.
20. The system of claim 15 , wherein:
the first type of data includes an accelerometer signal in an X direction;
the second type of data includes an accelerometer signal in a Y direction;
the target set of data further includes a fourth dimension for a third type of data; and
the third type of data includes an accelerometer signal in a Z direction.