 Hello supply chain designers and welcome to our first announced session. I'm Sergio Caballero, the core fleet And I'm very pleased to have Dr. Chris Cappi here with me. Welcome Chris Dr. Chris Capplis is the executive director at the MIT Center for the Center for Transportation and Logistics and he's also the instructor of our cars Okay, before getting started. Let me give you some statistics about the cars We have more than 20,300 students already enrolled in this class The median age of our students is 30 years Wow, that's older than normal, isn't it? Yeah, I think so. I think so. 24 percent of our students are women and our students represent more than 180 countries all over the world The total country with most students are the US with 19 percent, India 11 percent and Brazil with 3 percent Okay, this is the third time that we are running SC2x and the course has been redesigned So now the course we have eight weeks of content Two of weeks one meter and one final nothing new for those who have already taken SC0 and SC1 1x Yeah, so it's following the identical pattern of SC1x, which the majority of you probably already taken And then we have what the problem was before they were never originally designed to have a mid-term or a final And we didn't want to have a homework great assignment overlapping with a final That's why we put in that breather week and week five and then also be week 11 And so we always have a week before the test for you to catch up And I think the structure works well, but we'd love to get your feedback Excellent, so let me address the agenda for the for the for this session I'm going to start asking you some some questions and general questions about the about the course Then I will read some of your questions guys. So please now at the time Write your any question that you that you might have I'm going to read later Then we'll have a breakout session And we already posted the infractions in the In the week with two material And you will have between 10 to 15 minutes to come up with some some solution to that a particular problem posted Great after that we will get we'll read your answer and if time allows We'll again have some a Q&A session. Okay, so let it get started So, uh, Chris this first time now it's open to all the students So let me ask you this question What an audience student that is taking a c2x should start thinking to become a 35 students Yeah, so there's uh So the underlying or the overarching reason why we created these courses Mainly pushed by president Raphael Reif and the director of the digital learning center, which is at Professor Sanjay Sarma is to educate the world for free So everything will always be there for free the videos the practice problems the quick questions And some assessment graded questions as well as a mid-term and the final So it'll always be there for free to educate However, the credentialization is something different that takes more work And that's if you're verified you'll get credentialization. So if you just want to Educate yourself pick something up then fine. The audit is fine and it'll always be there for you However, we found that in addition to being credentialized there are other benefits when you become a verified student And becoming verified requires two things one There's a charge of $150 per course and two you have to have the id verified You have to show id and have a picture taking themselves so we can validate that you're the person actually doing the work So other benefits from this besides having a credential that you can use on linkedin your resume and elsewhere It's also being accepted by other universities to include mit and other universities around the world Um, also we found that if you're a verified student you tend to do better. You tend to stay more engaged There's a phrase if you have skin in the game So it seems like if you pay a little bit of money, you're more engaged Verified students typically do much better than audit students. They stay longer. They do more of the work Also, what we try to do to encourage verified students We have a section each week that's called the supplemental materials for micro master students because if you're verified We assume you're on the path for the micro masters credential. And so what we put in this Separate area that's only accessible by people who are verified are supplemental materials Now these aren't things that uh, you have to do in order to pass the course Because all that's free for the audit students as well These are supplemental things like feces that have been done up here that are relevant to that week's topics additional problems Sometimes we'll have interviews or videos But there are things to kind of round out and give some a little extra value to those students who are actually Supporting the efforts as well So you get the credentialization you get access to additional materials You also get access to hangouts like this after this hangout Only the verified students will be allowed to come in and again We're just trying to incentivize people to become more verified for a couple reasons One is we want to become sustainable financially so we can keep doing this because someone has to pay our salaries But the second is it makes you a better student if you're verified you'll do better And so that's why we're trying to encourage people to become verified Whether you whether you want to go through the whole micro masters credential and come here for a master's degree or just want this one course It's a better way to take the course Excellent. Thank you. So I'm nearly a micro master, but something that Surprised me was the results that we get from x1x 80 percent of the students that have completely Completed the course are verified students I think this is the data that supports what you were your claim. Yeah, okay. So we have the liberal ready two weeks of content All of it was was about the labor design, right? But what should the students expect for the rest of the course? Yeah, so because of the redesign we didn't have an overarching Video to show you where it's going you can actually look at the syllabus and get that But let me give you a rough idea Everything with a supply chain design fc2x is all about the design of the three major flows Right, there's the physical flow the information flow on the financial flow We start with the physical flow and for the physical flow We our primary tool is mix-injured linear programming for supply chain network design It's the standard approach used by every company every consultancy out there What we've done in the last part I think we do less than two we talk about some advanced topics Right covered those and so different ways that you can also use that tool for other other purposes or how to enhance that Because as we all know mix-injured linear programs are deterministic Right, and so you can't handle uncertainty very well, although we give you some workarounds how to do robust optimization and do different runs But that's the physical flow Next we'll move in to some of the information flow and we're going to start from the production information Going to the customer and that's because the tools you use to plan your production schedule Are the same exact tools that you use to your supply chain network design In fact, the only difference is you add a time element remember for a strategic network design It's usually a single period and it's fixed if I do multiple periods Where inventory I need to know the inventory position at each point in time I'm a Essentially creating a production planning schedule which we'll talk about and that's what we talk about in weeks Three you can continue this on further to understand the information flow from your company To your customers by doing aggregate planning and aggregate planning is something we'll cover on week four again using A mix-in-jured linear program to see what's the effect of running a promotion where it changes the demand What's the effect of changing my capacity at my plant or my labor hours? so the same Methodology mix-in-jured linear program large-scale optimization is used to map and coordinate the information flow For manufacturing to distribution to your customers So we do the physical flow Then we start doing the information flow And we're going to have a midterm after that we continue the information flow And we actually go backwards a little bit we talk about how you can interact with your supplier and the reason why we shifted that is because at um in sc2x In week six when you take your midterm you're at the exact midpoint of the entire program You've taken zero sc1x and half of sc2x and up to that point everything's been very quantitative and mathematical everything's a model Starting with the second half of sc2x into sc3x and forex It gets a little more qualitative and so there's a lot more Softer things that are being considered the models are still there, but there are other things that are being considered So what we start afterwards after the midterm is how do you deal with your suppliers? Sourcing those kind of things and then we'll go into a two and a half to three week segment on finance specifically looking how supply chain ties in to financial decisions will because when you work in supply chain You're talking about physical metrics when you talk in the language of the cfo the chief financial officer You need to talk in terms of dollars And so we'll talk about that translation You kind of a mini course on supply chain finance and then we'll finish off with organizational design So essentially what we're doing is physical flow design of physical flow design of information flow Design of the financial flow and then we finish with organizational design How should you structure your supply chain organization? And we'll capstone that with a case study which is one of probably one of the first ones That we've done here in this course so far, but that kind of gives you the big picture everything's about the design of the three flows Okay, thank you So many of our students please have already taken sc0 and sc1 and us and most of them are familiar with sass or ample So my question to you Chris is why is this cool stuff using primarily the three sheets? Yeah, so Um We did sc2x This is its third run and we did this prior to doing sc0x when we decided to offer more software solutions There's no easy way to do this and so originally when we did these courses sc1 Zero two and three We assumed we're going to stay in excel because the nice thing about spread sheets excel Libre offers google sheets is it can do a little bit of everything pretty well But as we all know it doesn't do certain things larger scale very well It has it has issues and limitations And so when we redid sc0x for the first time in the fall and have run again Now in the middle of running it again We gave the option to students to use a wide range of software packages And you can still use all of those packages if you have to do a linear program And I show how to do it in excel you can do it in ample You can do it in r if you want you can do it in cplex if you want the answer for any of the problems will be the same If you no matter what package you use assume it's an okay package But I kept the spreadsheet stuff in there Just because it's simple it's another way to look at it because we didn't show you how to do it in excel or Libre office prior to this we took that out of zero So I assume it's just a compliment and the thing that I like about doing in a spreadsheet that you don't have in sass Or an ample you don't see the constraint matrix. I love seeing the constraint matrix And so to me whenever I think of a problem I think of it visually As a constraint matrix and so you can think of it where you're going to have an identity matrix What your binary variables look like to me it makes total sense when I can have the columns and so I'm very comfortable seeing it in that view because that's how I grew up not doing it in excel But being able to map out the the matrices However, if you can use any tool you want so don't think that you have to use Excel for this you can use sass or ample for any of these and you've already done all of these in sc0x You've done transportation problems. You've done trans shipment problems. You've done mix-ins or linear programs So it's all the same you can you reuse some of those models or those approaches to solve every one of these problems Okay, thank you and what do you prefer increase? What do you usually so what do you say to solve the practical parts? Yeah, I haven't if it's something small. I'll slap it in excel In 30 seconds, I'll do that really quick. I've done that for a bunch of things I I created a little quick optimization to assign students to something and I did it in like 20 minutes excel Because it was easy for something larger scale like we have a project We tend to use cplex, which is something a little more It has a huge license fate and also it solves bigger problems It's more production level and so for larger projects We'll use cplex for me just messing around with something. I'll just use excel But we've used other things you can also have add-ins for excel that can make it much bigger There's an add-in called Am I what's it called open source? What's the new? Yeah for solver. What was it called? Yeah, but it doesn't work on the Mac So there you have to look around but there are some open source ones. Sadly nothing works on a Mac That's one of the biggest problems we've had. I love max, but they have issues But there are other solvers out there So I I don't do a ton of coding myself these days anymore But you should answer that question. What's your favorite optimization tool? So in my case, I use a lot of python for all the modeling Okay, and to solve the problem the solver that I tend to use is is groovy But again, you will have you'll need the license to that to that. Yeah, okay So but the important thing is to do that You're python surrounding it. So if you're going to do this in production You're not going to enter the things in by hand like we typically do for these kind of simple problems You'll have code surrounding it that will take an input matrix and output a matrix Okay, okay, so we have a some questions from the students. All right, so let me read the first one So how to balance the science and art effort a model can be built to very complex and require many hours to be to be built Okay, so how do you balance these science and art about the yes, they would sign Yes, you need to balance it. Yeah, I mean, it's it's an open-ended question So I think one of the challenges that you face when you have an optimization It's easier now than it used to be when I was doing a lot of this work encoding myself in The late 80s It was hard to do a lot of this now. It is so easy to have an optimization model. It's massive. So You have this yearning to try to create these really complex Detailed strategic models. The problem is you can't execute them And we'll talk more about that when we talk about combinatorial options in a week Seven and eight I think with procurement where you can have this really intricate solution But it can't be executed. So it's a judgment call The art the science can be taught very quickly. We can teach you the science And we'll talk about that during your question the art of how to when is too much too much and how to apply that That's a challenge and that usually requires judgment for the context To know to understand when you're going too far similar when you guys did like regression You don't always try to go for that highest r squared value Right, you want to find something that gives you a reasonable answer and unfortunately when students are doing this for the first time They have zero context. So they don't know how to have that That's why it's helpful to have people who've been seasoned and they can start tempering when the judgment comes I can't teach you art. Um, you'd have to experience that in my opinion Okay, excellent. I think it's now time to go through the breakout rooms. Okay. Can I can I read the question then? Okay, I want to put my glasses on because I'm getting old okay, so What we asked and they have this already Yes, okay So the big thing I wanted to do for this is At this question that was just asked is a perfect example Balancing this art in science because we know the math works But the math works under a lot of assumptions that don't always apply So the question I wanted to ask you guys is take a look at the real world now and over the last 12 months We've seen a lot of surprising political events election results. Um, you know brexit They voted voted to leave the eu in uh, june of last year Uh, donald trump was elected president that was totally unexpected. He went on a platform building a wall and um trying to reduce the amount of global trade and uh, to be honest his opponent Hilary clinton was also Leaning against global trade too. So it was another interesting twist and france the second round presidential elections in two weeks And you couldn't ask for more polar Candidates you also have uncertainty happening in turkey. You have uncertainty happening in china right now There's a lot of political uncertainty So the question is if I give you a hypothetical company that's operating in europe And you currently have manufacturing in a handful of countries and you distribute throughout And you're being asked to come up with a potential supply chain design plan for the next 10 years What would you do to try to Represent or capture some of this uncertainty or to guard against it because you don't know what's going to happen in 10 years Or over the course of 10 years But companies have to make decisions that they can't change quickly And so what I want you guys to do is in your groups come up with like three things no more than three Bullet point them so we can read it. Don't write essays. Just make it short. Think think like you're using twitter think like the president Right come up with something uh short. But what would you do for the network design problem? Okay, and what are the three things you would do differently and let's talk about it because this is a real problem And I'll talk about a real case that's happening now So what do they have 10 minutes? So they will have a 15 minutes 15 minutes. Well, well 15 minutes There we go 10 and 5 All right, see you 15 see you never be