R-PMS System Use Cases : Model . KDOT PMS Process and Data Flow Overview : Pool
Choreography Sub-Process - PMS Modeling Processes link
Properties |
Procedure |
Steps | |||
1. Lock the existing PMS Highway Network | |||
2. Query the data elements based upon a specific lock date (called a Snapshot by the AECom system). | |||
3. Calculate the current state of the system using the query data (snapshot) | |||
3.1. Define the segment for the modeling | |||
3.1.1. AECom utilizes 1 mile segments | |||
3.1.2. Alternative processes can utilize 1/200th mile segments | |||
3.2. Define the Organization level (normalization level) | |||
3.2.1. AECom utilizes the Road Category | |||
3.2.2. Alternative methods can utilize individual road segments | |||
4. Normalize query data into comparatives and aggregate values such as: | |||
4.1. Road Categories | |||
4.2. Distress States | |||
5. Estimate the remaining life of each segment | |||
6. Model the road surface conditions for future years | |||
6.1. AECom procedures | |||
6.1.1. Create an eventual "Steady State" final condition as a target | |||
6.1.2. Transition optimization procedures based upon goals | |||
6.1.2.1. Maximize Benefit within specific cost cap | |||
6.1.2.2. Minimize Costs with constrained by Road Categories | |||
6.1.3. Use the Optimization results to create an Action plan | |||
6.1.4. Rank the Action plan | |||
6.1.5. Identify Candidate Projects | |||
6.1.6. Match the Candidate Projects to Segments | |||
6.1.7. Generate Candidate Project Scopes | |||
6.2. ML/AI postulated procedures (1/200 mile segment) | |||
6.2.1. Utilize the historical scan data segment | |||
6.2.2. Determine historical wear patterns per segment | |||
6.2.3. Acquire modeled future Traffic conditions | |||
6.2.4. Acquire modeled future weather conditions | |||
6.2.5. Acquire modeled future resource availabilities | |||
6.2.6. Acquire modeled future Economic conditions | |||
6.2.7. Blend it all together utilizing industry standard ML/AI procedures such as: | |||
6.2.7.1. Trained Neural Nets using Historical Scan Data | |||
6.2.7.2. Deep Learning Neural Nets | |||
6.2.7.3. Multivariate Regression Analysis | |||
6.2.7.4. Dimensionality Reduction e.g. Primary Component Analysis | |||
6.2.7.5. Genetic Algorithms | |||
6.2.7.6. Statistical Algorithms | |||
6.2.7.7. etc. | |||
6.2.8. Generate Action Plans on aggregated segments (Projects) | |||
6.2.9. Rank Action Plans | |||
6.2.10. Generate Candidate Lists | |||
6.2.11. Match Candidate Lists to Segments | |||
6.2.12. Generate Candidate Project Scopes |
Relationships Summary |
Name | Begin | End | |||
|
|
|
|||
|
|
|
|||
|
|
|
|||
|
|
|
|||
|
|
|
Relationships Detail |
Name | Value | ||
Name | 1R Candidate Project LIst | ||
From |
|
||
To |
|
Name | Value | ||||||||
Name | 3R Comparative Candidate List | ||||||||
From |
|
||||||||
To |
|
||||||||
Transit From |
|
Name | Value | ||
Name | |||
From |
|
||
To |
|
||
Id | BP221.BP188 | ||
Direction | None |
Name | Value | ||
Name | |||
From |
|
||
To |
|
||
Id | BP221.BP146 | ||
Condition Type | None | ||
Immediate | Unspecified | ||
Quantity | 1 | ||
Timeout | 0 |
Name | Value | ||
Name | |||
From |
|
||
To |
|
Sub Diagrams |
Name | Description | ||
|
Details of the Proposed PMS optimization modeling activities showing data products and activity/action states. |
References |
File Name | Description | ||
|
|||
N/A | |||
N/A | |||
N/A | |||
N/A | |||
N/A | |||
|
|||
|
|||
|
|||
|
Appears In |
Diagram | |||
|