System and method for dependency and root cause discovery
The embodiments described herein describe a computerized system and method for retrieving and processing data to provide dependency and root cause information for a process. The computerized system and method include receiving historic data of the process, detecting temporal dependency or precedence tasks in the process from the historic data, generating a historic dependency graph, aggregating the historic dependency graph into a probabilistic precedence network (PPN), pruning the PPN, and presenting results to a user.
1. A computerized method for retrieving and processing data to provide dependency and root cause information for a process, comprising:
receiving data corresponding to a historic performance of the process;
detecting temporal dependencies and precedence tasks in the process from the historic data, wherein the temporal dependency is a dependency among tasks that are related through time constraint;
generating historic dependency data;
aggregating the historic dependency data into a probabilistic precedence network (PPN), the PPN including a complete set of all possible temporal dependency relationships reflected in the historic data;
pruning the PPN by removing dependencies that occur a minimum number of times based on the historic dependency data, wherein the minimum number of times is a predefined minimum threshold value;
identifying, from tasks that maintained a dependency in the pruned PPN, tasks that have had at least one of a delay or variance above a defined threshold based on the historic data;
determining, from the tasks that have had at least one of a delay or variance above a defined threshold, tasks that have had a failure or a delay from one or more of the following: rework, repair, part shortage, and operator request for parts;
identifying the tasks that have had a failure or a delay from one or more of the following: rework, repair, part shortage, and operator request for parts as a root cause for delay; and
presenting results to a user.
2. The method according to claim 1 , wherein the temporal dependency or the precedence tasks are represented as a precedence network in the form of directed acyclic graph (DAG).
3. The method according to claim 1 , wherein the dependency data includes, for the entire process, all required tasks to manufacture a product.
4. The method according to claim 1 , wherein pruning the PPN further comprises filtering out less likely dependencies of the process and keeping robust dependencies of the process.
5. The method according to claim 4 , wherein a particular dependency will be kept in the PPN if a probability of the particular dependency exceeds a threshold.
6. The method according to claim 4 , wherein the robust dependencies reflect tasks that have mandatory precedence relationships that impact the entire process.
7. A system for retrieving and processing data to provide dependency and root cause information for a process, comprising:
a memory area; and
a processor programmed to:
collect data corresponding to a historic performance of the process;
detect temporal dependencies and precedence tasks in the process from the historic data, wherein the temporal dependency is a dependency among tasks that are related through time constraint;
generate historic dependency data;
aggregate the historic dependency data into a probabilistic precedence network (PPN), the PPN including a complete set of all possible temporal dependency relationships reflected in the historic data;
prune the PPN by removing dependencies that occur a minimum number of times based on the historic dependency data, wherein the minimum number of times is a predefined minimum threshold value;
identify, from tasks that maintained a dependency in the pruned PPN, tasks that have had at least one of a delay or variance above a defined threshold based on the historic data;
determine, from the tasks that have had at least one of a delay or variance above a defined threshold, tasks that have had a failure or a delay from one or more of the following: rework, repair, part shortage, and operator request for parts;
identify the tasks that have had a failure or a delay from one or more of the following: rework, repair, part shortage, and operator request for parts as a root cause for delay; and
present results to a user.
8. The system according to claim 7 , wherein the temporal dependency or the precedence tasks are represented as a precedence network in the form of directed acyclic graph (DAG).
9. The system according to claim 7 , wherein the dependency data includes, for the entire process, all required tasks to manufacture a product.
10. The system according to claim 7 , wherein pruning the PPN comprises filtering out less likely dependencies of the process and keeping robust dependencies of the process.
11. The system according to claim 10 , wherein a particular dependency will be kept in the PPN if a probability of the particular dependency exceeds a threshold.
12. The system according to claim 10 , wherein the robust dependencies reflect tasks that have mandatory precedence relationships that impact the entire process.