Jim Connett, on July 07, 2021, 09:21 AM
E10 Unmasked – Closing the Loop
In December of 2018, we published a four-part series about the SEMI E10 standard. Since that time, we’ve received excellent questions highlighting some of the troublesome nuances of correctly applying E10. Here are some questions and answers we think are beneficial to the entire community. As more questions arrive, we’ll continue to respond directly as well as summarize the questions and answers in this format. Do you have a question? Let us know at email@example.com.
Q1: What use-case exists that underscores the importance of the E10 data in manufacturing?
A1: One of the service-level agreement metrics provided on capital manufacturing equipment may be a metric called “UPTIME” – this is the amount of time a tool’s manufacturer will guarantee that the tool will be in STANDBY, PRODUCTIVE, or ENGINEERING throughout a rolling pre-defined period (usually 4-weeks or 13-weeks). Warranty events are often based on these metrics, and when the tool you’ve purchased falls below these guaranteed levels, the equipment manufacturer needs to address the performance of the tool at no cost to the purchaser. Tracking E10 data provides this data for the UPTIME calculation which will underscore your request for warranty work. In the end, this alone can save you thousands of dollars.
Furthermore, collecting the time accrued by equipment in each of the E10 states produces all of the variable values needed to calculate OEE metrics like utilization, availability, cycle time, mean time between failure, mean time to failure, etc. Over time, E10 begins to tell a story that helps you maximize tool performance and target resources toward resolving issues with its performance.
Q2: How should we handle the time gap between the point when the material arrives at a tool and the tool starting the process? Sometimes, tools need to adjust their temperatures or pressures based on the chosen recipe for the material. How do we account for these points of precondition prior to processing?
A2: This one is tough to answer. It depends on the data you wish to collect and the triggering event that would cause the state change from STANDBY to PRODUCTIVE. In most cases, the MES’s move-in or track-in event signals the end of the current state (very likely STANDBY) and the beginning of the PRODUCTIVE state as the material begins to process. Likewise, the move-out or track-out event signals the end of the PRODUCTIVE state and the beginning of the STANDBY state. So, if the move-in transaction occurs before the equipment is ready to receive the first unit of raw material into its processing sequence, then the preparatory time between move-in and the tool’s START event will be counted as PRODUCTIVE. Technically, this is accurate because the tool has control of the raw material and is configured to process this raw material. However, long pre- and post-processing equipment requirements may skew the actual PRODUCTIVE data (for example, furnace tubes that take 30–45 minutes to temp-up and pump down, and the same amount of time to temp-down and return to atmosphere at the end of processing).
Properly configured OEE data collection systems can account for these pre- and post-processing times so that the PRODUCTIVE time reflects actual “chemicals-on-raw-material” processing. But that’s totally up to the company and the value of the data being collected.
Q3: Imagine a scenario where a scheduled preventive maintenance (PM) procedure puts a tool down to a SCHEDULED DOWN state. At the end of the maintenance tasks, a qualification process is required to ensure the tool can receive production material following the PM. Does the E10 state change when the tool is being qualified after a PM?
A3: If the qualification is a required step in the PM process, and if a passing qualification is a precondition to marking the PM as DONE, then we suggest the equipment should stay in the SCHEDULED DOWN E10 state. If the qualification fails, then the PM is likely the cause (a chamber wasn’t purged before the qual, the temperature didn’t stabilize, or a cleanroom wipe was left in the load-lock – yes, I’ve seen that one before!) and the PM steps will need to be verified or redone.
If, however, the qualification is NOT required, yet an Engineer is concerned about the tool’s state after the PM, then it would be appropriate for the SCHEDULED DOWN state to end when the PM ends, and the tool should transition to either ENGINEERING or maybe UNSCHEDULED DOWN during the qualification process.
To take this example to an extreme conclusion, if it is determined that the qualification after preventive maintenance is failing due to a pressure sensor in a transfer chamber (for example) and you have no stock of this specific part and need to order it, then this situation would be a good candidate for a transition from SCHEDULED DOWN to an UNSCHEDULED DOWN state…or even a NONSCHEDULED DOWN state because the problem is now an external event (no parts available).
In general, if the task or step is required as part of the PM, we advise the tool remain in a SCHEDULED DOWN state. Any additional tasks following the completed PM should cause the tool to transition to an UNSCHEDULED DOWN state.