Optima KPI methodology

Template ID	Template Name	Provider	Description
1	→ Runs in Delay - Provider: Optima Public Transport	optima-put	Percentage of Public Transport runs with a delay. The KPI is calculated to return a percentage number [0-100] corresponding to the total number of runs that have a delay greater than zero with respect to the total running runs.
2	→ Aggregate Delay - Provider: Optima Public Transport	optima-put	Sum of the delays, measured in minutes, of all active runs referring to the next stops. The KPI is calculated to return the overall delay of the Public Transport system as a sum of all delays in reaching the next stops of all running runs.
3	→ Average Delay - Provider: Optima Public Transport	optima-put	Average delay, measured in minutes, referring to the next stops for all the active runs. The KPI is calculated to return the average delay of the Public Transport system as an average of all delays in reaching the next stops of all running runs.
4	→ Maximum Delay - Provider: Optima Public Transport	optima-put	Maximum delay, measured in minutes, referring to the next stops for all the active runs. The KPI is calculated to return the maximum delay of the Public Transport system as the maximum delay in reaching the next stops of all running runs.
5	→ Aggregate Delay on Stop - Provider: Optima Public Transport	optima-put	Sum of the delays of all the active runs referring either to a specified stop or to all of them. The KPI is calculated to return the aggregation of all delays for a given stop or for all active runs passing through it.
6	→ Average Delay on Stop - Provider: Optima Public Transport	optima-put	Average of the delays of all the active runs referring either to a specified stop or to all of them. The KPI is calculated to return the average of all delays, for a given stop or for all active runs passing through it.
7	→ Maximum Delay on Stop - Provider: Optima Public Transport	optima-put	Maximum delay of all the active runs referring either to a specified stop or to all of them. The KPI is calculated to return the maximum delay of all the active runs referring to specified stops.
8	→ Percentage Monitored Runs - Provider: Optima Public Transport	optima-put	Percentage of monitored vehicles with respect to the total of running ones. The KPI is calculated to return the share of vehicles which have provided their position in the latest PuT vehicle real-time feed, with respect to the total of vehicles currently circulating (live plus by schedule).
9	→ Average User Speed - Provider: Short-Term Forecast	Short-Term Forecast	Average user speed as simulated by the Short-Term Forecast (STF) engine all over the selected network. The KPI is calculated by handling the average speed of all users as a weighted average of the simulated speeds, with weights set to the corresponding number of simulated vehicles. If the number of simulated vehicles is missing in the network selection, the average of the model speeds is automatically set. The KPI is calculated for the specified forecast horizon and considering only equivalent vehicles.
10	→ Average Network Speed - Provider: Short-Term Forecast	Short-Term Forecast	Average speed on the network as simulated by the Short-Term Forecast (STF) engine all over the selected network. The KPI is calculated by handling the average speed of all links as a weighted average of the simulated speeds, which are computed as a ratio between the travel times spent by users to travel the links and the link lengths, with weights set to the amount of simulated flow on the links. If the number of simulated vehicles is 0 in the network selection, the weighted average of the link speeds (freeflow, possibly modified by events) is automatically set, with weights set to the link length. The KPI is calculated for the specified forecast horizon and considering only equivalent vehicles.
11	→ Number of Vehicles on the Network - Provider: Short-Term Forecast	Short-Term Forecast	Number of vehicles in the network.
12	→ Network Vehicle Output - Provider: Short-Term Forecast	Short-Term Forecast	Network vehicle output, measured as the average flow (veh/h) which leaves the network during the interval. The KPI is calculated as the average flow (veh/h) which leaves the network and enters any of the destination connectors during the interval.
13	→ Network Vehicle Input - Provider: Short-Term Forecast	Short-Term Forecast	Network vehicle input, measured as the average flow (veh/h) which enters the network during the interval. The KPI is calculated as the average flow (veh/h) which enters the network and leaves any of the origin connectors during the interval.
14	→ Total Travel Time - Provider: Short-Term Forecast	Short-Term Forecast	Total travel time, measured in hours, calculated as cumulative travel time on any set of streets. The KPI is calculated by taking, for each single street, the product between: The number of vehicles (input flow on the street). The travel time of the vehicles in the interval of results for the selected forecast distance.
15	→ Production - Provider: Short-Term Forecast	Short-Term Forecast	Vehicle production for the selected set of links. It is computed as the entry flow on the specified links during the selected interval multiplied by the link lengths.
16	→ Harmonizer: Average Speed Detections - Provider: Optima Harmonizer	optima-harmonizer	Average speed detections, measured in Km/h. The KPI is calculated considering, for the selected streets, the average of the harmonized speeds in the specified interval. The harmonized values of speed considered are the values with the starting instant of validity falling within the interval.
17	→ Harmonizer: Average Flow Detections - Provider: Optima Harmonizer	optima-harmonizer	Average flow detections, measured in Veh/h. The KPI is calculated by considering, for the selected streets, the average of the harmonized flows in the specified interval. The harmonized values of flow considered are the values with the starting instant of validity falling within the interval.
18	→ Harmonizer: Count Speed Detections - Provider: Optima Harmonizer	optima-harmonizer	Number of vehicles in the network.
19	→ Harmonizer: Count Flow Detections - Provider: Optima Harmonizer	optima-harmonizer	Number of active flow detections. The KPI is calculated by considering, for the selected streets, the number of the harmonized measurements in the specified interval. For example, considering that the Harmonizer is run every 5 minutes, if the interval is set to 15 minutes, you can count up to 3 harmonized measurements for every street . The harmonized values of flow considered are the values with the starting instant of validity falling within the interval.
20	→ Model Forecast Quality from Base (Flow) - Provider: Short-Term Forecast	Short-Term Forecast	Quality indicator of the forecast flow vs. the offline model.
21	→ Model Forecast Quality (Flow) - Provider: Short-Term Forecast	Short-Term Forecast	Quality indicator of the model forecast vs. the measured speed.
23	→ KPI templates	machine-learning-forecast-quality	Quality indicator of the Machine Learning forecast vs the measured flow.
24	→ Average User Speed - Provider: Optima Planning	optima-planning	Average user speed as simulated by the Optima Planning engine all over the selected network. The KPI is calculated by handling the average speed of all users as a weighted average of the simulated speeds, with weights set to the corresponding number of simulated vehicles. If the number of simulated vehicles is missing in the network selection, the average of the model speeds is automatically set.
25	→ Average Network Speed - Provider: Optima Planning	optima-planning	Average speed on the network as simulated by the Optima Planning engine all over the selected network. The KPI is calculated by handling the average speed of all links as a weighted average of the simulated speeds, which is computed as a ratio between the travel times spent by users to travel the links and the link lengths, with weights set to the amount of the simulated flow on the links. If the number of simulated vehicles is 0 in the network selection, it is set to the weighted average of the link speeds (freeflow, possibly modified by events), with weights set to the link length.
26	→ Number of Vehicles on the Network - Provider: Optima Planning	optima-planning	Number of vehicles in the network, measured as the simulated average number of vehicles on all the links of the network. The KPI is calculated as the simulated average number of vehicles of all the links of the network. This equals to taking a snapshot of the network during an instant of the reference interval and to counting the number of vehicles in the snapshot; the average number comes from averaging the numbers for every instant within the interval.
27	→ Network Vehicle Output - Provider: Optima Planning	optima-planning	Network vehicle output, measured as the average flow (veh/h) which leaves the network during the interval. The KPI is calculated as the average flow (veh/h) which leaves the network and enters any of the destination connectors during the interval.
28	→ Network Vehicle Input - Provider: Optima Planning	optima-planning	Network vehicle input, measured as the average flow (veh/h) which enters the network during the interval. The KPI is calculated as the average flow (veh/h) which enters the network and leaves any of the origin connectors during the interval.
29	→ Total Travel Time - Provider: Optima Planning	optima-planning	Total travel time, measured in hours, calculated as cumulative travel time on any set of links. The KPI is calculated by taking, for each link, the product between: The number of vehicles (input flow on the link). The travel time spent by the vehicles.
30	→ Production - Provider: Optima Planning	optima-planning	Vehicle production for the selected set of links. It is computed as the entry flow on the specified links during the selected interval multiplied by the link lengths.
31	→ PTV Optima Micro: Travel Times - Provider: Optima Micro	Optima Micro	Travel time, measured in minutes, associated with a path or area. The KPI is calculated along a path or summed up from all the links of an area according to Optima Micro results. It is only applicable within an Optima Micro subnetwork.
32	→ PTV Optima Micro: Total Vehicle Times - Provider: Optima Micro	Optima Micro	Total vehicle time, measured in hours, spent on a path or area. The KPI is calculated considering the total time spent by all vehicles along a path or on all links of an area, according to Optima Micro results. It is only applicable within an Optima Micro subnetwork.
33	→ Path Travel Time - Provider: Short-Term Forecast	Short-Term Forecast	Average travel time, measured in minutes, spent by vehicles on the selected path of the network. The KPI is calculated as the sum of the travel times of every link along the path, as estimated by PTV Optima. If the space-time trajectory of the vehicle along the path falls inside a result time interval, the used travel time is interpolated between the instantaneous travel times computed at the start and end times of the simulation intervals.
34	→ Path Travel Time - Provider: Optima Planning	optima-planning	Average travel time, measured in minutes, spent by vehicles on the selected path of the network. The KPI is calculated as the sum of the travel times of every link along the path, as estimated by PTV Optima. If the space-time trajectory of the vehicle along the path falls inside a result time interval, the used travel time is interpolated between the instantaneous travel times computed at the start and end times of the simulation intervals.
35	→ ML Forecast: Congestion Index - Provider: Machine Learning Forecast	machine-learning-forecast	The Congestion index is a percentage. It can vary from 0 to infinity. The lower the index, the better. Index=0 means that the network speed coincides on every street with its free flow speed. The KPI is calculated considering the % of additional time for traveling through the network streets.
36	→ ML Forecast: Traffic Jam - Provider: Machine Learning Forecast	machine-learning-forecast	Total length of queues, measured in [Km]. The KPI is calculated as the total length of streets whose average speed is lower than 50% of their free flow speed.
37	→ ML Forecast: Network Congestion Share - Provider: Machine Learning Forecast	machine-learning-forecast	Network congestion share is a percentage. It can vary from 0 to 100%. The lower the index, the better. Index=0 means that the network speed coincides on every street with its free flow speed. The KPI is calculated as the % of streets whose average speed is lower than 50% of their free flow speed.
38	→ Harmonizer: Congestion Index - Provider: Optima Harmonizer	optima-harmonizer	Congestion index is a percentage. It can vary from 0 to infinity. The lower the index, the better. Index=0 means that the network speed coincides on every street with its free flow speed. The KPI is calculated considering the % of additional time for traveling through the network streets.
39	→ Harmonizer: Traffic Jam - Provider: Optima Harmonizer	optima-harmonizer	Total length of queues, measured in [Km]. The KPI is calculated as the total length of streets whose average speed is lower than 50% of their free flow speed.
40	→ Harmonizer: Network Congestion Share - Provider: Optima Harmonizer	optima-harmonizer	The network congestion share is a percentage. It can vary from 0 to 100%. The lower the index, the better. Index=0 means that the network speed coincides on every street with its free flow speed. The KPI is calculated as the % of streets whose average speed is lower than 50% of their free flow speed.
41	→ Percentage of Links below Error Threshold (Flow) - Provider: Short-Term Forecast	Short-Term Forecast	Percentage of links for which the Model Forecast Quality (Flow) is below the threshold specified in the template's attributes.
42	→ Model Forecast Quality from Base (Speed) - Provider: Short-Term Forecast	Short-Term Forecast	It is an indicator of how close the measurement of the current speed is to the average estimate made for the current day-type. The KPI is calculated as the relative error between the harmonized traffic measurements (speed) of the current speed and the estimate for the offline simulation of the current day-type.
43	→ Model Forecast Quality (Speed) - Provider: Short-Term Forecast	Short-Term Forecast	It is an indicator of how accurate the forecast made for the current speed in the past at varying forecast distances is. The KPI is calculated as the relative error between the harmonized traffic measurements (speed) of speeds at the current moment and past forecasts.
44	→ Machine Learning Forecast Quality (Speed) - Provider: Optima Machine Learning Forecast Quality	machine-learning-forecast-quality	Quality indicator of the Machine Learning forecast vs the measured speed.
45	→ Percentage of Links below Error Threshold (Speed) - Provider: Short-Term Forecast	Short-Term Forecast	Percentage of links for which the Model Forecast Quality (Speed) is below the threshold specified in the template's attributes.
46	→ ML Forecast: Percentage of Streets below Error Threshold (Speed) - Provider: Optima Machine Learning Forecast	machine-learning-forecast	The percentage of streets below the error threshold, measured as the relative error of the forecast speed (at the given forecast distance) versus the measured speed.
47	→ ML Forecast: Percentage of Streets below Error Threshold (Flow) - Provider: Optima Machine Learning Forecast	machine-learning-forecast	The percentage of streets below the error threshold, measured as the GEH of the forecast flow (at the given forecast distance) versus the measured flow.

1

→ Runs in Delay - Provider: Optima Public Transport

optima-put

Percentage of Public Transport runs with a delay.

The KPI is calculated to return a percentage number [0-100] corresponding to the total number of runs that have a delay greater than zero with respect to the total running runs.

2

→ Aggregate Delay - Provider: Optima Public Transport

optima-put

Sum of the delays, measured in minutes, of all active runs referring to the next stops.

The KPI is calculated to return the overall delay of the Public Transport system as a sum of all delays in reaching the next stops of all running runs.

3

→ Average Delay - Provider: Optima Public Transport

optima-put

Average delay, measured in minutes, referring to the next stops for all the active runs.

The KPI is calculated to return the average delay of the Public Transport system as an average of all delays in reaching the next stops of all running runs.

4

→ Maximum Delay - Provider: Optima Public Transport

optima-put

Maximum delay, measured in minutes, referring to the next stops for all the active runs.

The KPI is calculated to return the maximum delay of the Public Transport system as the maximum delay in reaching the next stops of all running runs.

5

→ Aggregate Delay on Stop - Provider: Optima Public Transport

optima-put

Sum of the delays of all the active runs referring either to a specified stop or to all of them.

The KPI is calculated to return the aggregation of all delays for a given stop or for all active runs passing through it.

6

→ Average Delay on Stop - Provider: Optima Public Transport

optima-put

Average of the delays of all the active runs referring either to a specified stop or to all of them.

The KPI is calculated to return the average of all delays, for a given stop or for all active runs passing through it.

7

→ Maximum Delay on Stop - Provider: Optima Public Transport

optima-put

Maximum delay of all the active runs referring either to a specified stop or to all of them.

The KPI is calculated to return the maximum delay of all the active runs referring to specified stops.

8

→ Percentage Monitored Runs - Provider: Optima Public Transport

optima-put

Percentage of monitored vehicles with respect to the total of running ones.

The KPI is calculated to return the share of vehicles which have provided their position in the latest PuT vehicle real-time feed, with respect to the total of vehicles currently circulating (live plus by schedule).

9

→ Average User Speed - Provider: Short-Term Forecast

Short-Term Forecast

Average user speed as simulated by the Short-Term Forecast (STF) engine all over the selected network.

The KPI is calculated by handling the average speed of all users as a weighted average of the simulated speeds, with weights set to the corresponding number of simulated vehicles.

If the number of simulated vehicles is missing in the network selection, the average of the model speeds is automatically set.

The KPI is calculated for the specified forecast horizon and considering only equivalent vehicles.

10

→ Average Network Speed - Provider: Short-Term Forecast

Short-Term Forecast

Average speed on the network as simulated by the Short-Term Forecast (STF) engine all over the selected network.

The KPI is calculated by handling the average speed of all links as a weighted average of the simulated speeds, which are computed as a ratio between the travel times spent by users to travel the links and the link lengths, with weights set to the amount of simulated flow on the links.

If the number of simulated vehicles is 0 in the network selection, the weighted average of the link speeds (freeflow, possibly modified by events) is automatically set, with weights set to the link length.

The KPI is calculated for the specified forecast horizon and considering only equivalent vehicles.

11

→ Number of Vehicles on the Network - Provider: Short-Term Forecast

Short-Term Forecast

Number of vehicles in the network.

12

→ Network Vehicle Output - Provider: Short-Term Forecast

Short-Term Forecast

Network vehicle output, measured as the average flow (veh/h) which leaves the network during the interval.

The KPI is calculated as the average flow (veh/h) which leaves the network and enters any of the destination connectors during the interval.

13

→ Network Vehicle Input - Provider: Short-Term Forecast

Short-Term Forecast

Network vehicle input, measured as the average flow (veh/h) which enters the network during the interval.

The KPI is calculated as the average flow (veh/h) which enters the network and leaves any of the origin connectors during the interval.

14

→ Total Travel Time - Provider: Short-Term Forecast

Short-Term Forecast

Total travel time, measured in hours, calculated as cumulative travel time on any set of streets.

The KPI is calculated by taking, for each single street, the product between:

The number of vehicles (input flow on the street).
The travel time of the vehicles in the interval of results for the selected forecast distance.

15

→ Production - Provider: Short-Term Forecast

Short-Term Forecast

Vehicle production for the selected set of links. It is computed as the entry flow on the specified links during the selected interval multiplied by the link lengths.

16

→ Harmonizer: Average Speed Detections - Provider: Optima Harmonizer

optima-harmonizer

Average speed detections, measured in Km/h.

The KPI is calculated considering, for the selected streets, the average of the harmonized speeds in the specified interval.

The harmonized values of speed considered are the values with the starting instant of validity falling within the interval.

17

→ Harmonizer: Average Flow Detections - Provider: Optima Harmonizer

optima-harmonizer

Average flow detections, measured in Veh/h.

The KPI is calculated by considering, for the selected streets, the average of the harmonized flows in the specified interval.

The harmonized values of flow considered are the values with the starting instant of validity falling within the interval.

18

→ Harmonizer: Count Speed Detections - Provider: Optima Harmonizer

optima-harmonizer

Number of vehicles in the network.

19

→ Harmonizer: Count Flow Detections - Provider: Optima Harmonizer

optima-harmonizer

Number of active flow detections.

The KPI is calculated by considering, for the selected streets, the number of the harmonized measurements in the specified interval.

For example, considering that the Harmonizer is run every 5 minutes, if the interval is set to 15 minutes, you can count up to 3 harmonized measurements for every street .

The harmonized values of flow considered are the values with the starting instant of validity falling within the interval.

20

→ Model Forecast Quality from Base (Flow) - Provider: Short-Term Forecast

Short-Term Forecast

Quality indicator of the forecast flow vs. the offline model.

21

→ Model Forecast Quality (Flow) - Provider: Short-Term Forecast

Short-Term Forecast

Quality indicator of the model forecast vs. the measured speed.

23

→ KPI templates

machine-learning-forecast-quality

Quality indicator of the Machine Learning forecast vs the measured flow.

24

→ Average User Speed - Provider: Optima Planning

optima-planning

Average user speed as simulated by the Optima Planning engine all over the selected network.

The KPI is calculated by handling the average speed of all users as a weighted average of the simulated speeds, with weights set to the corresponding number of simulated vehicles.

If the number of simulated vehicles is missing in the network selection, the average of the model speeds is automatically set.

25

→ Average Network Speed - Provider: Optima Planning

optima-planning

Average speed on the network as simulated by the Optima Planning engine all over the selected network.

The KPI is calculated by handling the average speed of all links as a weighted average of the simulated speeds, which is computed as a ratio between the travel times spent by users to travel the links and the link lengths, with weights set to the amount of the simulated flow on the links.

If the number of simulated vehicles is 0 in the network selection, it is set to the weighted average of the link speeds (freeflow, possibly modified by events), with weights set to the link length.

26

→ Number of Vehicles on the Network - Provider: Optima Planning

optima-planning

Number of vehicles in the network, measured as the simulated average number of vehicles on all the links of the network.

The KPI is calculated as the simulated average number of vehicles of all the links of the network. This equals to taking a snapshot of the network during an instant of the reference interval and to counting the number of vehicles in the snapshot; the average number comes from averaging the numbers for every instant within the interval.

27

→ Network Vehicle Output - Provider: Optima Planning

optima-planning

Network vehicle output, measured as the average flow (veh/h) which leaves the network during the interval.

The KPI is calculated as the average flow (veh/h) which leaves the network and enters any of the destination connectors during the interval.

28

→ Network Vehicle Input - Provider: Optima Planning

optima-planning

Network vehicle input, measured as the average flow (veh/h) which enters the network during the interval.

The KPI is calculated as the average flow (veh/h) which enters the network and leaves any of the origin connectors during the interval.

29

→ Total Travel Time - Provider: Optima Planning

optima-planning

Total travel time, measured in hours, calculated as cumulative travel time on any set of links.

The KPI is calculated by taking, for each link, the product between:

The number of vehicles (input flow on the link).
The travel time spent by the vehicles.

30

→ Production - Provider: Optima Planning

optima-planning

Vehicle production for the selected set of links. It is computed as the entry flow on the specified links during the selected interval multiplied by the link lengths.

31

→ PTV Optima Micro: Travel Times - Provider: Optima Micro

Optima Micro

Travel time, measured in minutes, associated with a path or area.

The KPI is calculated along a path or summed up from all the links of an area according to Optima Micro results.

It is only applicable within an Optima Micro subnetwork.

32

→ PTV Optima Micro: Total Vehicle Times - Provider: Optima Micro

Optima Micro

Total vehicle time, measured in hours, spent on a path or area.

The KPI is calculated considering the total time spent by all vehicles along a path or on all links of an area, according to Optima Micro results.

It is only applicable within an Optima Micro subnetwork.

33

→ Path Travel Time - Provider: Short-Term Forecast

Short-Term Forecast

Average travel time, measured in minutes, spent by vehicles on the selected path of the network.

The KPI is calculated as the sum of the travel times of every link along the path, as estimated by PTV Optima.

If the space-time trajectory of the vehicle along the path falls inside a result time interval, the used travel time is interpolated between the instantaneous travel times computed at the start and end times of the simulation intervals.

34

→ Path Travel Time - Provider: Optima Planning

optima-planning

Average travel time, measured in minutes, spent by vehicles on the selected path of the network.

The KPI is calculated as the sum of the travel times of every link along the path, as estimated by PTV Optima.

If the space-time trajectory of the vehicle along the path falls inside a result time interval, the used travel time is interpolated between the instantaneous travel times computed at the start and end times of the simulation intervals.

35

→ ML Forecast: Congestion Index - Provider: Machine Learning Forecast

machine-learning-forecast

The Congestion index is a percentage.

It can vary from 0 to infinity. The lower the index, the better.

Index=0 means that the network speed coincides on every street with its free flow speed.

The KPI is calculated considering the % of additional time for traveling through the network streets.

36

→ ML Forecast: Traffic Jam - Provider: Machine Learning Forecast

machine-learning-forecast

Total length of queues, measured in [Km].

The KPI is calculated as the total length of streets whose average speed is lower than 50% of their free flow speed.

37

→ ML Forecast: Network Congestion Share - Provider: Machine Learning Forecast

machine-learning-forecast

Network congestion share is a percentage.

It can vary from 0 to 100%. The lower the index, the better.

Index=0 means that the network speed coincides on every street with its free flow speed.

The KPI is calculated as the % of streets whose average speed is lower than 50% of their free flow speed.

38

→ Harmonizer: Congestion Index - Provider: Optima Harmonizer

optima-harmonizer

Congestion index is a percentage.

It can vary from 0 to infinity. The lower the index, the better.

Index=0 means that the network speed coincides on every street with its free flow speed.

The KPI is calculated considering the % of additional time for traveling through the network streets.

39

→ Harmonizer: Traffic Jam - Provider: Optima Harmonizer

optima-harmonizer

Total length of queues, measured in [Km].

The KPI is calculated as the total length of streets whose average speed is lower than 50% of their free flow speed.

40

→ Harmonizer: Network Congestion Share - Provider: Optima Harmonizer

optima-harmonizer

The network congestion share is a percentage.

It can vary from 0 to 100%. The lower the index, the better.

Index=0 means that the network speed coincides on every street with its free flow speed.

The KPI is calculated as the % of streets whose average speed is lower than 50% of their free flow speed.

41

→ Percentage of Links below Error Threshold (Flow) - Provider: Short-Term Forecast

Short-Term Forecast

Percentage of links for which the Model Forecast Quality (Flow) is below the threshold specified in the template's attributes.

42

→ Model Forecast Quality from Base (Speed) - Provider: Short-Term Forecast

Short-Term Forecast

It is an indicator of how close the measurement of the current speed is to the average estimate made for the current day-type.

The KPI is calculated as the relative error between the harmonized traffic measurements (speed) of the current speed and the estimate for the offline simulation of the current day-type.

43

→ Model Forecast Quality (Speed) - Provider: Short-Term Forecast

Short-Term Forecast

It is an indicator of how accurate the forecast made for the current speed in the past at varying forecast distances is.

The KPI is calculated as the relative error between the harmonized traffic measurements (speed) of speeds at the current moment and past forecasts.

44

→ Machine Learning Forecast Quality (Speed) - Provider: Optima Machine Learning Forecast Quality

machine-learning-forecast-quality

Quality indicator of the Machine Learning forecast vs the measured speed.

45

→ Percentage of Links below Error Threshold (Speed) - Provider: Short-Term Forecast

Short-Term Forecast

Percentage of links for which the Model Forecast Quality (Speed) is below the threshold specified in the template's attributes.

46

→ ML Forecast: Percentage of Streets below Error Threshold (Speed) - Provider: Optima Machine Learning Forecast

machine-learning-forecast

The percentage of streets below the error threshold, measured as the relative error of the forecast speed (at the given forecast distance) versus the measured speed.

47

→ ML Forecast: Percentage of Streets below Error Threshold (Flow) - Provider: Optima Machine Learning Forecast

machine-learning-forecast

The percentage of streets below the error threshold, measured as the GEH of the forecast flow (at the given forecast distance) versus the measured flow.