-
Notifications
You must be signed in to change notification settings - Fork 35
Reliability
Revised : 15 Nov 2021
Beginning r4133, we have implemented new reliability modeling to supplement with how the maintenance and malfunctions are being handled in the simulation for each item resource (aka part).
3 Basic metrics of Reliability :
| Metrics | Unit |
|---|---|
| Mean Time Between Failure (MTBF) | (Earth) hours |
| Failure Rate | % |
| Reliability | % |
By definition, MTBF, an industry standard that gives the average time between failures for a component or a Part.
Incidentally, MTBF for a given part/component is also the adjusted value of the inverse of the Failure Rate.
At the beginning, each part is given a Reliability of 99.999% and a standard Failure Rate of 3%.
For simplicity sake, in each pre-defined malfunction, we also derive a composite metric called the Probability of Failure from the 3 metrics above.
Below is a snapshot of the command prompt log screen :
00-Adir-07:662.377 (Warning) Malfunction : Class A Fire - incident #3 - an Emergency repair work order was requested.
00-Adir-07:662.377 (Warning) Malfunction : Class A Fire - incident #3 - a General repair work order was set up.
00-Adir-07:662.377 (Warning) Malfunction : Class A Fire - incident #3 - the repair requires fire extinguisher (quantity: 2).
00-Adir-07:662.377 (Warning) Malfunction : Class A Fire - incident #3 - the repair requires fiberglass cloth (quantity: 2).
00-Adir-07:662.377 (Warning) MalfunctionManager : [ 1.69° N 19.74° E] Dune Runner has Class A Fire as reported by Anaisha Kappor. Cause : Human Factors.
00-Adir-07:662.377 (Warning) MalfunctionManager : --- Part : fiberglass cloth ---
00-Adir-07:662.377 (Warning) MalfunctionManager : (1). Reliability : 99.577 % --> 99.48 %
00-Adir-07:662.377 (Warning) MalfunctionManager : (2). Failure Rate : 0.063 % --> 0.078 %
00-Adir-07:662.377 (Warning) MalfunctionManager : (3). MTBF : 1414.5 hr --> 1150.5 hr
00-Adir-07:662.377 (Warning) MalfunctionManager : --- Malfunction : Class A Fire ---
00-Adir-07:662.377 (Warning) MalfunctionManager : (4). Probability : 3.0 % --> 1.539 %
00-Adir-07:662.377 (Warning) MalfunctionManager : [ 1.69° N 19.74° E] Anaisha Kappor had an accident.
00-Adir-07:662.377 (Warning) MalfunctionManager : [ 1.69° N 19.74° E] A Type-I accident occurred in Dune Runner in Dune Runner.
00-Adir-07:663.600 (Info) Malfunction : Air Leak - incident #2 - Emergency repair work initiated by Pari Shan.
00-Adir-07:678.364 (Info) Malfunction : Class A Fire - incident #3 - Emergency repair work initiated by Pari Shan.
When an incident/malfunction occurs over a part, these 3 metrics (as well as the probability of failure of that malfunction) will be recomputed and updated dynamically.
The MalfunctionManager class is responsible for doing the update whenever a malfunction occurs.
Obviously the reliability % gets better and the MTBF of a part would improve on each sol having no malfunction occurs.
Note: regardless, new reliability data is re-computed daily.
The MalfunctionFactory class is responsible for storing the reliability, failure rate and MTBF of each Part.
See https://github.com/mars-sim/mars-sim/issues/507 for adding the ability to showcase the reliability data for each part.