Production Performance Dashboards - Calculations Explained
D
Dan Miner
started a topic
12 months ago
Each tile represents a single production metric for 1 to many selectable stations / lines. The header bar at the top (green in the sample below) will be either green or red. If any of the tiles are red then the bar at the top would be red.
For each production metric, hourly means the transaction time stamps within current time minus 60 minutes, daily means current time minus 24 hours, weekly means the transaction time stamps within current time minus - 7 days, monthly means current time minus 30 days; yearly means current time minus 365 days, YTD means from Jan 1 through the current time.
Users can select which time bucket (hourly, daily, weekly, monthly, yearly, YTD) they want to include.
Users can set the number of places past the decimal that they want to include.
Users can adjust the header text per tile.
The production metrics would be the following:
Process Lead Time - Attempts to find the average duration of a USN production based on the earliest sign on and final sign off. How long was it in the plant, dept, line, station based on the selection of the stations.
The user will select the stations that they want to consider (expected that they will select the first and last stations on the line in most cases).
The user should be able to set the target as a float value per tile. Green is any value less than or equal to the target. Single target applies to all time horizons. The user can set the units that lead time will be calculated against (minutes, hours, days, weeks, months)
For each selected time horizon, select all USNs which have a completed operation sign off for all selected stations.
For each USN, subtract the oldest operation sign on time from the newest operation sign off time. A selection option exists to 'Remove scheduled downtime' based on the work schedule (For example, if the Sign on occurs at 3pm on Friday and sign off at 10am on Monday. They are scheduled down from 4pm on Friday till 8am on Monday. It should result in 3 hours of lead time). Then take the average of the lead time values.
First Pass Yield - Attempts to determine a ratio of units which pass through a process without a defect.
The user will select the stations that they want to consider (expected that they will select a single station on the line, but they may consider multiple. If the user wants to see the first pass yield for multiple stations in separate tiles, then they need to create multiple FPY tiles. If they want to see an aggregate FPY, then they will select multiple stations in a single tile.
The user should be able to set the target as a float value per tile. Green is any value greater than or equal to the target. Single target applies to all time horizons.
The user can set if they want to display the time horizon results as a percent or a count (e.g. 67% vs 2/3). This is always a unitless calculation.
The user can set which types of defects are considered in the FPY calculation within the tile for the included station(s)
Work Step Tests - RequiredInSpec would need to be set to false in order for the work step test results to differ from the spec.
Test Failures - At least one test failure with a status not equal to cancelled
Roadblocks - At least one roadblock with a status not equal to cancelled (No need to allow for specifying roadblock types)
Activity Skips - At least one skip was used (implies a failure)
Torque Tools - At least one test torque value falls outside the min/max torque values
Hot Issues - At last one hot issue filed
NOTE:
Dimension Data - Not Included. The 'Dimension Data' inputs which tie back to the item master custom fields do not have a pass/fail associated to them. This currently allows for free form text. The users are expected to input a roadblock or similar if there is a dimensional failure.
PFEP - Not included. No eKitting or eKanban 'failures' would be included.
For the numerator, find the number of distinct USNs which are signed off at all stations with zero defects.
For the denominator, find the number of distinct USNs which signed on to at least one station with a defect count greater than zero OR are signed off at all stations.
Throughput - Attempts to find rate at which USNs are moving through the included stations.
The user will select the stations that they want to consider (expected that they will select the last station on the line in most cases).
The user should be able to set the target as a float value per tile. Green is any value greater than or equal to the target. Single target applies to all time horizons. The user can set the units of the throughput calculation (USNs per minute, hour, day, month)
For each selected time horizon, select all USNs which have a completed operation sign off for all selected stations.
Include all complete time buckets where the bin size is current time minus x (60 minutes, 24 hours, 7 days, 30 days, 365 days). Count the number of sign offs in each bucket then average based on the count of buckets. Include buckets with zero sign offs. Add a selection option to 'Exclude buckets with zero sign offs and scheduled downtime'.
Dan Miner
Each tile represents a single production metric for 1 to many selectable stations / lines. The header bar at the top (green in the sample below) will be either green or red. If any of the tiles are red then the bar at the top would be red.
For each production metric, hourly means the transaction time stamps within current time minus 60 minutes, daily means current time minus 24 hours, weekly means the transaction time stamps within current time minus - 7 days, monthly means current time minus 30 days; yearly means current time minus 365 days, YTD means from Jan 1 through the current time.
Users can select which time bucket (hourly, daily, weekly, monthly, yearly, YTD) they want to include.
Users can set the number of places past the decimal that they want to include.
Users can adjust the header text per tile.
The production metrics would be the following:
Process Lead Time - Attempts to find the average duration of a USN production based on the earliest sign on and final sign off. How long was it in the plant, dept, line, station based on the selection of the stations.
The user will select the stations that they want to consider (expected that they will select the first and last stations on the line in most cases).
The user should be able to set the target as a float value per tile. Green is any value less than or equal to the target. Single target applies to all time horizons. The user can set the units that lead time will be calculated against (minutes, hours, days, weeks, months)
For each selected time horizon, select all USNs which have a completed operation sign off for all selected stations.
For each USN, subtract the oldest operation sign on time from the newest operation sign off time. A selection option exists to 'Remove scheduled downtime' based on the work schedule (For example, if the Sign on occurs at 3pm on Friday and sign off at 10am on Monday. They are scheduled down from 4pm on Friday till 8am on Monday. It should result in 3 hours of lead time). Then take the average of the lead time values.
First Pass Yield - Attempts to determine a ratio of units which pass through a process without a defect.
The user will select the stations that they want to consider (expected that they will select a single station on the line, but they may consider multiple. If the user wants to see the first pass yield for multiple stations in separate tiles, then they need to create multiple FPY tiles. If they want to see an aggregate FPY, then they will select multiple stations in a single tile.
The user should be able to set the target as a float value per tile. Green is any value greater than or equal to the target. Single target applies to all time horizons.
The user can set if they want to display the time horizon results as a percent or a count (e.g. 67% vs 2/3). This is always a unitless calculation.
The user can set which types of defects are considered in the FPY calculation within the tile for the included station(s)
Work Step Tests - RequiredInSpec would need to be set to false in order for the work step test results to differ from the spec.
Test Failures - At least one test failure with a status not equal to cancelled
Roadblocks - At least one roadblock with a status not equal to cancelled (No need to allow for specifying roadblock types)
Activity Skips - At least one skip was used (implies a failure)
Torque Tools - At least one test torque value falls outside the min/max torque values
Hot Issues - At last one hot issue filed
NOTE:
Dimension Data - Not Included. The 'Dimension Data' inputs which tie back to the item master custom fields do not have a pass/fail associated to them. This currently allows for free form text. The users are expected to input a roadblock or similar if there is a dimensional failure.
PFEP - Not included. No eKitting or eKanban 'failures' would be included.
For the numerator, find the number of distinct USNs which are signed off at all stations with zero defects.
For the denominator, find the number of distinct USNs which signed on to at least one station with a defect count greater than zero OR are signed off at all stations.
Throughput - Attempts to find rate at which USNs are moving through the included stations.
The user will select the stations that they want to consider (expected that they will select the last station on the line in most cases).
The user should be able to set the target as a float value per tile. Green is any value greater than or equal to the target. Single target applies to all time horizons. The user can set the units of the throughput calculation (USNs per minute, hour, day, month)
For each selected time horizon, select all USNs which have a completed operation sign off for all selected stations.
Include all complete time buckets where the bin size is current time minus x (60 minutes, 24 hours, 7 days, 30 days, 365 days). Count the number of sign offs in each bucket then average based on the count of buckets. Include buckets with zero sign offs. Add a selection option to 'Exclude buckets with zero sign offs and scheduled downtime'.