That was nice. We have went through the modeling process that we indeed actually went through when we were working with this company. The company gave us this problem. We used the OR modeling tools to do the survey, to do the research. Now it's time to present you with the result. So result is presented here. This is the optimal solution we obtained by using the Gurobi optimizer. With our mathematical model, after we input all the information that the company provides, we are able to deliver this particular solution. Pretty much, we still want all the facilities. But for each facility, the company gives us two options. Either we keep the current setting, which means a large-scale, or it's possible that we, in some sense cut down or shut down that facility, but to give a small office with just one engineer there. The company somehow is suggesting that for a facility, maybe we may choose not to rent it at all, but we choose to rent a small office around it. If that's the case, the small office may still host one engineer. Then that one engineer may still serve some customers, but at least there are some important customers may be served by a close facility. After the optimization, we see that there are five locations that you don't need a large facility. For Danshui, Hsinchu, Tainan, Yilan, and Taitung, you actually don't need a large building. You just need to have one Building with just one engineer. That engineer is going to serve those most important customers around it. For most other customers, for example, for those other customers in Danshui, Nangang office is going to take care of that. For many customers in Tainan, you will use Chiayi and Kaohsiung to serve those customers in Tainan, and so on and so on. The optimizer tells you that this is the best way to save office rents. Maybe it's because that currently office rent is too high and gasoline fees not so high. For engineers allocation, you may also see that there are some huge offices that need a lot of engineers. For those small-scale offices, because each of office may host at most one engineer, so you only have one engineer at small offices. You also have the number of customers assigned, all the customers, and then you also have the number of services that each facility need to take care. Of course, that's a feasible solution. We may also do have some interesting interpretations. First, we do see that downsizing to small facilities is helpful, at least for five facilities. Probably that's really because 20 years ago, office rent is not so high, but today, it's too high. Probably we should not have so many huge facilities. Instead, we should have some small facilities around Taiwan. Also, it's still nice to have large facilities in both Kaohsiung and Pingtung. That's actually a very interesting findings because previously you may think this two places are so close to each other. What's the point of having two facilities together? Well, the reason may be that the office rent for Pingtung is so cheap. It doesn't really matter whether you have the Kaohsiung facility, you always want the Pingtung facility. Then somehow the optimizer states that, "Well, why don't you do that? You host one facility at Pingtung and host seven people there. " Those seven engineers, if you put all of them in Kaohsiung, first, the utility fees are higher in Kaohsiung per engineer. In the second, there are still some customers here, maybe in Hanchueng or somewhere. If you travel from Kaohsiung to Hanchueng, that's just too far. If you have a Pingtung facility, then that cost you saved for having a closer facility is good enough to cover the cheap facility cost. If you don't have an optimizer, maybe you don't think in that way. But if you put all the numbers there, the optimizer tells you this is the optimal solution. There are still some discussions if you want to implement the solution in practice. First, we may see that there are some benefits, because you are actually building fewer large facilities. You're going to save some money, and that that about seven million or eight million something is around seven percent of the total cost per year. I would say it's a huge amount of money. But of course, there are some potential detriments. You need to subsidize engineers for reallocation. That costs some money. You need to spend something to maintain customer relationship. Because there are some customers size that the responsible engineer would be changed and you probably need to take care of those customers. Finally, it may be uneasy for you to find just one single engineer to take care of one region. Because you need that engineer to be able to do a lot of maintenance skills that may be difficult. But at least, now you will have a way to quantify the potential benefits. You may evaluate whether this is large enough for you to take care of all these potential detriments. If this number is very small, probably you will give up the downsizing plan. But if the benefit, if the cost-saving is very huge, then you really want to do that, even if you have those potential, detriments. With the motto, now it's easier for you to make decisions, not just based on intuition, not just based on experience. Now you have more numbers, more concrete evidence for you to make decisions either in the current environment for today or in future environments.
