We’re doing a Round Two of our favourite segment of these editorials, Tyler White Tries to Teach You Everything That He Learned in His Semester Two Classes (not including electives since you won’t necessarily do MECHENG 211 and I’m too lazy to teach that)
ENGGEN 204: Managing Design and Communication
I like to think of this paper in three segments: communication, systems thinking, and general career advice and tidbits.
Communication
Most of this is done in tutorials. This includes writing reports, learning *basic* Excel skills, delivering oral presentations, and creating posters.
From the lectures, I can offer three morsels of knowledge:
- 70% of face-to-face communication is body language. Body language provides feedback on whether you are communicating well or if you need to change your communication.
- Email and texting are great for confirming or seeking basic information, but not for critical or complex information. There is no feedback on your communication. Also, text is not as nuanced and fine-tuned as your voice and your physical presence.
- Communication is what the other party understands. Communication is not what you say.
Systems Thinking
At Part II, Systems Thinking is the ability to take complex and poorly defined problems and simplifying them down. Simplification allows us to create solutions easier. We practiced this by learning about engineering ethics and sustainability. Engineering ethics boils down into “do what your granny would want you to do”. Sustainability boils down into “don’t fuck over the future’s environments, economies, or societies”.
Career Tidbits
- Engineers must combine their Hard Skills (theory, knowledge) with their Soft Skills (Interpersonal skills, commercial skills, managerial skills)
- It is not enough to attend lectures and count that as ‘participation’. Participation requires us to think about what we’re learning.
- Students believe that there are job-ready upon graduation. Employers think otherwise.
- You are hired as a profit centre for someone else.
ENGSCI 255: Modelling in Operations Research
This is the optimisation paper that everyone uses to label Engineering Science as ‘That Optimisation Degree’. This paper has three sections: Linear Programming (not at all like computer programming), Networks and Project Management, and Queueing.
What exactly is optimisation? Optimisation is simply maximising or minimising something. This is achieved by using different levers (or ‘decision variables’). An example of an optimisation problem might be minimising the amount of time you spend studying for your exams. The different decision variables are the amount of time you spend studying for each exam.
Linear Programming (not at all like computer programming)
The easiest way to minimise time studying in the described example is to spend no time studying at all. Alas, life is not as simple as this. For starters, you probably have a GPA target that you want to hit. You need time to sleep and eat so that you don’t go insane. You also have a finite amount of time before the exams come around. These are called ‘constraints’, and must be mathematically described as part of the problem. It is these constraints that require you to be strategic as to how you allocate your time. With Operations Research, you learn the tools that help you be as strategic as possible with your time in this problem.
Let’s say you give this problem to Excel to solve. The software will give you the best amounts of time to spend on each exam, given the rules that you gave Excel (i.e. I promised my mum I’d spend at least 10 hours on my CHEMMAT 121 exam, I only have a total of 100 hours to study for all exams).
But! Imagine this! You realize that you underestimated the benefits of studying for ENGGEN 131 – does that change your solution? Or maybe you forgot that you agreed to help your friend move houses during the weekend? Does this change how you spend your time? What if you could copy a friend’s notes for some of the exams? You learn how to deal with situational changes like these in ENGSCI 255.
Of course, ENGSCI 255 can be used to answer so many more questions. Where should I build my next shoe factory? When should I roster my staff of nurses and doctors to minimise patient waiting time? These problems can be solved with linear programs.
Networks and Project Management
Networks are a series of nodes or points that are interconnected. Nodes can be locations, people, computer servers – anything that can be connected to one another.
If you want to move from one node to another in the cheapest/shortest way possible, operations research can help you in a variety of ways. If you want to connect the nodes together into one complete system in the cheapest manner, operations research can help you. If you want to travel along the cheapest circuit and visit every node in the network, operations research will do a bloody good job of this. These networks are being optimised to minimise (or maximise) something, be it time, money, or distance.
As for the project management bit, think of a project as a series of tasks that must be done to achieve something. Examples might include organising events, creating products, pulling off heists, etc.
Project management allows us to understand which tasks must be finished on time and which tasks can fall behind schedule. Project management tells us how long a project should take. Project management can tell us the best way to spend money to shorten the duration of the project.
Why are Networks and Project Management clumped together under the same section? The tasks in each project are connected in a way: I can’t do tasks B or C without doing task A first. Each task is a node, and each task is connected depending on which tasks are dependent on other tasks.
Queueing
A queue is a system where people wait to be served by someone. Examples include customers waiting to make an order at McDonalds, cars reaching a toll booth, or students waiting to see their lecturer at office hours.
Queues are wasteful. People who are waiting in queues don’t like wasting time. Some people might leave the queue. This is wasteful on the server since they’ve lost customers. Long queues suck. Period. We could fix this by hiring an infinite number of servers. This way, a server is always free to serve a customer. Customers would never have to wait.
Alas, hiring an infinite number of servers is also wasteful on our budget (and impossible). While long queues waste time, eliminating the queue entirely is an unnecessary waste of money. With knowledge of queueing models, we can balance the two and help the system operate as quickly as possible at minimal cost.
How does one attain knowledge of queueing models in ENGSCI 255? The first component of any system with a queue is that people will usually arrive at random times. Randomness implies probability. Probability refers to averages and to the probability distributions that you learned in high school and in ENGSCI 111.
The second component of a queueing system is that the servers will take some time to serve each customer. The service time is also typically random, so we must fall back on averages and probability distributions.
Finally, we must know how many servers are operating. A system with zero servers is beyond hope. A system with one server might work. A system with two servers might be even better. Whether or not we have enough servers depends on how many arriving customers the system is under and how quickly each server can operate.
Once we have enough servers to avoid a catastrophe, we can begin to answer questions such as: How busy will the servers typically be? How many people might there typically be in the system? What happens if we restrict the number of people who can wait in queue? Is it worth hiring another server?
ENGSCI 263: Engineering Science Design I
This is a strange course. This course is made of two independent mini-classes that run at the same time and never affect each other at all. The theme for each class is the same. The classes in Semester One were about understanding the world around us and how we could describe it using mathematics or mechanics or computers. ENGSCI 263 is the next step: taking this knowledge, and using it to make things.
ENGSCI 263 has a Design strand and a Modelling strand. The Design strand focuses on general design skills and principles. The Modelling strand specialises on making mathematical models in the proper fashion (not like the ghetto-ass modelling in ENGSCI 111 or ENGSCI 211).
The Design Strand
The Design strand consists entirely of two different projects, plus a few labs to bring you up to speed with the technology that we use. The first project involves prototyping a mechanical device with Excel. If you are like Me Before Semester Two, your Excel skills are limited to using Excel as a table to put numbers into. After ENGSCI 263, you will know 80-90% of the useful Excel formulae that are used to do fancy things with the numbers. Also, you will know that Excel has a special computer programming language (yay!) called Visual Basic for Applications. The principles and uses of computer programming don’t change too much; it’s just that you can now write programs for Excel. This is awesome.
The second project involves using SolidWorks to speedily model the mechanical device. SolidWorks is like Creo, except that A) It’s a lot faster, B) You can use the programming in Excel to update the SolidWorks model, and C) People in the real world use SolidWorks. I’ve been informed that Creo isn’t used in the real world. Your ENGGEN 115 existence was a lie!
Asides from tangible skills, the Design strand emphasises the softer skills, such as approaching unwieldy problems and working in teams. A lot of what I learned here cannot be taught; it can only be truly learned via taking this class.
The Modelling Strand
I’ve already written about models in ENGSCI 211. Those are weak ass models. ENGSCI 263 is about real detailed and specific models for specific situations.
Quick recap: a mathematical model is something that accurately spits out numbers that represent something, provided you first give the model some other numbers that measure other things. This is the final product.
To reach the final product, we must first have a situation to model. ENGSCI 263 provides us with material most of us haven’t touched in high school or in Part I Engineering: Continuum Mechanics! If learning the delicacies of modelling wasn’t enough, students in ENGSCI 263 also learn to answer questions like: How does heat move through solid objects? Linearly? Radially? What temperatures occur at what locations in this medium? How fast is do fluids travel through different materials? How porous is this material? Considering everything else that EngSci students learn in other classes, you might be saying to yourself, “By gosh and by golly, these EngSci kids learn heaps!” My response: Yes, you’re right.
With our continuum mechanics ingredients all set, it’s time to mix them together with our modelling skills! There are three distinct things to think about: formulation, calibration, and uncertainty.
Formulation is a matter of combining science concepts together and using ninja-like derivation skills to create the model. Yes, you heard me: Derivation. The other specialisations get their equations on a plate. EngScis cook them up themselves. EngScis know where their food came from and how it was cooked. EngScis can state the assumptions that our models are built on, and how to defend them. Stating them is not enough (as I have found out in our assignments).
Calibration is about making sure we are using the right parameters, constants, and coefficients in our model. Chefs need to taste their food as they cook and season as necessary. Likewise, engineering scientists need to test their models against known data and modify parameters to get closer fits. This can range from “guess and check” to “seriously technical programming”.
Lastly, engineering scientists must acknowledge that there is uncertainty in their models. “All models are wrong, but some are useful” (George Box, 1976). I can’t say anything more because I haven’t watched the lectures yet. I’ve got plenty of time though.