Maintenance (MT) is not the most expensive part in company budget, but can save a couple thousand with proper planning. To do this, you need to invest as much - as you really need and no more.
I failed to find the exact dependence of the machines efficiency with MT, but by analogy with agents recruiting you can "draw" a map where you will visually see it. Future machines efficiency depends on:
- Previous machines efficiency.
- Normalized MT - MT, purified from the influence of machines load and number of shifts. In fact it will MT for one machine at 100% load (5 days) with 1 shift.
Normalized MT = MT in previous period / Number of shifts / (Machine hours / Total possible machine hours * Coefficient)
Coefficient = ratio of total possible machines hours for 7 days to the same number of machines hours for 5 days. In old version GMC simulator ratio is 1.4 to 1 shift and 1.3 to 2 or 3 shifts. In new version GMC simulator ratio is 1.37 for 1 shift and 1.27 for 2 and 3 shifts.
We get clean MT time based on 100% machines load. Also take into account changes in machines efficiency because of purchasing new machines. If you sold some machines, this data is not considered. Calculating the real machines efficiency after buying:
The real machines efficiency = (Machines efficiency in the next period * Machines in the next period - (Machines in the next period - Machines in the previous period) * 100) / Machines in the previous period
This map made from a large number of reports and well demonstrates that the formula of machines efficiency does not change between scenarios. Vertical - machines efficiency in the past period, the horizontal - normalized MT in the past period, the intersection - machines efficiency in the next period.
Hint - machines efficiency limit is achieved in the area of 90-92%, the cost to achieve higher efficiency is not cost-effective. In our calculation model we calculate optimal investments in MT for each scenario by every period. After trying all possible options (oldschool bruteforce) you will find the best MT plan. For example, in most cases, optimal plan for scenario 12C1 will look like:
- 1 period - 70
- 2 period - 50
- 3 period - 35
- 4 period - 25
- 5 period - 2