Object trail analytics
A method, a structure, and a computer system for object trail analytics. The exemplary embodiments may include obtaining time series data detailing an average speed of one or more roads within a traffic network at one or more times. The exemplary embodiments may further include extracting one or more features corresponding to the time series data, and generating one or more time series forecasting models based on the time series data and the one or more features. Additionally, the exemplary embodiments may include identifying a current location of a moving object within the traffic network, and predicting a speed of the moving object based on applying the one or more time series forecasting models to the current location.
1 . A computer-implemented method, comprising:
determining, by a system comprising a processor, time series data comprising respective average vehicle speeds on roads within a road network in respective time intervals of a group of time intervals, wherein the determining comprises monitoring, using distributed sensors of the road network, actual vehicle speeds of vehicles on the roads during the group of time intervals;
for each of the roads:
determining, using the distributed sensors of the road network, a respective group of features associated with the road at respective time points of a group of time points in a defined recurring time period corresponding to the time series data, and
generating and training a defined quantity of time series forecasting models based on the respective average vehicle speed on the road in the respective time intervals and the respective group of features at the respective time points, wherein the defined quantity is a function of a quantity of time intervals in the group of time intervals and a quantity of time points in the group of time points, wherein the training comprises training the time series forecasting models to predict vehicle speeds of vehicles on the road, next points of interest in the road network for the vehicles, and estimated times of arrival of the vehicles at the predicted next points of interest;
identifying, by the system, using the distributed sensors of the road network, a current location of a moving vehicle on a current road within the road network at a current time;
selecting, by the system, at least one time series forecasting model selected from the time series forecasting models that is associated with a road of the current location and a respective time point within a defined time window of the current time; and
predicting, by the system, using the at least one time series forecasting model and the current location, a vehicle speed of the moving vehicle on the current road.
2 . The computer-implemented method of claim 1 , further comprising:
identifying, by the system, using the distributed sensors of the road network, current features of the current road and one or more adjacent roads of the road network that satisfy a criterion associated with adjacency;
predicting, by the system, using the at least one time series forecasting model, a next point of interest in the road network for the moving vehicle based on the current road, the one or more adjacent roads, and the current features, wherein the predicting next point of interest comprises predicting vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at different future times based on the current features; and
estimating, by the system, using the at least one time series forecasting model, a time of arrival of the moving vehicle at the predicted next point of interest within the road network based on the predicted vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at the different future times.
3 . The computer-implemented method of claim 2 , further comprising:
determining, by the system, whether an elapsed time is greater than a difference between the current time and the estimated time of arrival; and
based on determining that the elapsed time is greater than the difference between the current time and the estimated time of arrival, identifying, by the system, the moving vehicle in an adjacent road of the one or more adjacent roads within a specified timeframe.
4 . The computer-implemented method of claim 3 ,
based on determining that the elapsed time is not greater than the difference between the current time and the estimated time of arrival, determining, by the system, that the moving vehicle has not yet reached the predicted next point of interest.
5 . The computer-implemented method of claim 1 , wherein the time series forecasting models correlate the respective average vehicle speeds on the roads with the respective groups of features associated with the roads in the respective time intervals.
6 . The computer-implemented method of claim 1 , wherein the respective groups of features comprise at least one of day, season, time, weather, or events.
7 . The computer-implemented method of claim 6 , wherein the events comprise at least one of road work, traffic accidents, or shows.
8 . A computer program product for object trail analytics, the computer program product comprising a non-transitory computer-readable media having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
determine time series data comprising respective average vehicle speeds on roads within a road network in respective time intervals of a group of time intervals, wherein the determining comprises monitoring, using distributed sensors of the road network, actual vehicle speeds of vehicles on the roads during the group of time intervals;
for each of the roads:
determining, using the distributed sensors of the road network, a respective group of features associated with the road at respective time points of a group of time points in a defined recurring time period corresponding to the time series data, and
generate and train a defined quantity of time series forecasting models based on the respective average vehicle speed on the road in the respective time intervals and the respective group of features at the respective time points, wherein the defined quantity is a function of a quantity of time intervals in the group of time intervals and a quantity of time points in the group of time points, wherein the training comprises training the time series forecasting models to predict vehicle speeds of vehicles on the road, next points of interest in the road network for the vehicles, and estimated times of arrival of the vehicles at the predicted next points of interest;
identify a current location of a moving vehicle on a current road within the road network at a current time;
select at least one time series forecasting model selected from the time series forecasting models that is associated with a road of the current location and a respective time point within a defined time window of the current time; and
predict, using the at least one time series forecasting model and the current location, a vehicle speed of the moving vehicle on the current road.
9 . The computer program product of claim 8 , wherein the program instructions are further executable by the processor to cause the processor to:
identify, using the distributed sensors of the road network, current features of the current road and one or more adjacent roads of the road network that satisfy a criterion associated with adjacency;
predict, using the at least one time series forecasting model, a next point of interest in the road network for the moving vehicle based on the current road, the one or more adjacent roads, and the current features, wherein the predicting next point of interest comprises predicting vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at different future times based on the current features; and
estimate, using the at least one time series forecasting model, a time of arrival of the moving vehicle at the predicted next point of interest within the road network based on the predicted vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at the different future times.
10 . The computer program product of claim 9 , wherein the program instructions are further executable by the processor to cause the processor to:
determine whether an elapsed time is greater than a difference between the current time and the estimated time of arrival; and
based on determining that the elapsed time is greater than the difference between the current time and the estimated time of arrival, identify the moving vehicle in an adjacent road of the one or more adjacent roads within a specified timeframe.
11 . The computer program product of claim 10 , wherein the program instructions are further executable by the processor to cause the processor to:
based on determining that the elapsed time is not greater than the difference between the current time and the estimated time of arrival, determine that the moving vehicle has not yet reached the predicted next point of interest.
12 . The computer program product of claim 8 , wherein the time series forecasting models correlate the respective average vehicle speeds on the roads with the respective groups of features associated with the roads in the respective time intervals.
13 . The computer program product of claim 8 , wherein the respective groups of features comprise at least one of day, season, time, weather, or events.
14 . The computer program product of claim 13 , wherein the events comprise at least one of road work, traffic accidents, or shows.
15 . A system, comprising:
a memory configured to store computer executable instructions; and
a processor configured to execute at least one of the computer executable instructions that:
determines time series data comprising respective average vehicle speeds on roads within a road network in respective time intervals of a group of time intervals, wherein the determining comprises monitoring, using distributed sensors of the road network, actual vehicle speeds of vehicles on the roads during the group of time intervals;
for each of the roads:
determines, using the distributed sensors of the road network, a respective group of features associated with the road at respective time points of a group of time points in a defined recurring time period corresponding to the time series data, and
generates and trains a defined quantity of time series forecasting models based on the respective average vehicle speed on the road in the respective time intervals and the respective group of features at the respective time points, wherein the defined quantity is a function of a quantity of time intervals in the group of time intervals and a quantity of time points in the group of time points, wherein the training comprises training the time series forecasting models to predict vehicle speeds of vehicles on the road, next points of interest in the road network for the vehicles, and estimated times of arrival of the vehicles at the predicted next points of interest;
identifies a current location of a moving vehicle on a current road within the road network at a current time;
selects at least one time series forecasting model selected from the time series forecasting models that is associated with a road of the current location and a respective time point within a defined time window of the current time; and
predicts, using the at least one time series forecasting model and the current location, a vehicle speed of the moving vehicle on the current road.
16 . The system of claim 15 , wherein the at least one of the computer executable instructions further:
identifies, using the distributed sensors of the road network, current features of the current road and one or more adjacent roads of the road network that satisfy a criterion associated with adjacency;
predicts, using the at least one time series forecasting model, a next point of interest in the road network for the moving vehicle based on the current road, the one or more adjacent roads, and the current features, wherein the predicting next point of interest comprises predicting vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at different future times based on the current features; and
estimates, using the at least one time series forecasting model, a time of arrival of the moving vehicle at the predicted next point of interest within the road network based on the predicted vehicle speeds of the moving vehicle on the current road and the one or more adjacent roads at the different future times.
17 . The system of claim 16 , wherein the at least one of the computer executable instructions further:
determines whether an elapsed time is greater than a difference between the current time and the estimated time of arrival; and
based on determining that the elapsed time is greater than the difference between the current time and the estimated time of arrival, identifies the moving vehicle in an adjacent road of the one or more adjacent roads within a specified timeframe.
18 . The system of claim 17 , wherein the at least one of the computer executable instructions further:
based on determining that the elapsed time is not greater than the difference between the current time and the estimated time of arrival, determines that the moving vehicle has not yet reached the predicted next point of interest.
19 . The system of claim 15 , wherein the time series forecasting models correlate the respective average vehicle speeds on the roads with the respective groups of features associated with the roads in the respective time intervals.
20 . The system of claim 15 , wherein the respective groups of features comprise at least one of day, season, time, weather, or events.