Why can can't we use $1-($lambda$/$mue$)$ for part a$.$ QdobQ, a yummy food restaurant, is very popular during lunchtime. The manager wants to perform a study on the service level of the restaurant. He hired an operations analyst to take this study. The findings are as follows. The counter of the restaurant can process only one customer at a time. During lunchtime, approximately there are $10$ customers in every $20$ minutes $($assuming this is a Poisson stream$).$ It takes on average $3$ minutes for the counter to help one customer $($assuming the service time follows exponential distribution$).$ The waiting line cannot exceed $25$ people. a$.$What is the probability the counter can relax during lunchtime? b$.$ What is the average amount of people waiting in the line? c$.$What is the expected total duration for a customer to purchase the food? d$.$What if QdobQopen another counter? Please recalculate $($a$),($b$),$ and $($c$).$