Robot Art Show

The goal of this project was to develop a working electronic device for the "robot art show" at the end of the year. (This was just a name for the informal presentation style — the constructed device did not have to be related to art or artwork.) The design process encompassed the wide variety of skills and concepts related to electronics, programming, and robotics in general. Because of this, we studied many of these subjects in class, including:

• Static electricity

• Power generation (Turbines)

• Electromagnetism (Motors)

• Circuit principles — Ohm's law, Kirchoff's laws, series, parallel, etc

• Programming and the Arduino framework

The project started to materialize once we got into circuits and programming. First, we made some basic series and parallel circuits with resistors and batteries. We used multimeters to verify Ohm's law (see the Core Concepts section) on these circuits and then began to make more complex circuits with light bulbs, LEDs, potentiometers and later capacitors, IC chips, and more. This culminated in the "Blinky Light Quiz" where we were tasked with blinking a light using a 555 timer chip. After this, we moved our focus to using Arduino boards to create more complex behavior. We used the Arduino C++ environment to create programs for a variety of experiments, including a digital piano, a memory game, and a guessing game. At this point, the class was ready to start working on the robot art show itself.

In addition, the lander mechanism has an additional servo which gives the lander rotational freedom, allowing the user to rotate the lander in order to change the angle of thrust. This input is provided by the joystick, which controls thrust (up) and rotation (left-right). Internally, this is then processed and accurate laws of physics are used to calculate the movement of the lander in "space". The goal of the game is to safely land the lander on the paper-mache lunar surface, and if this is achieved the player receives a celebration message on the LCD screen. If not — if they went too fast, or if the lander fell over on impact — then a failure message is shown. A video of the game running can be seen below.

Circuit Diagram and Completed Code

Core Concepts

Some of these concepts were not directly used in the final project, however they were studied in class and thus they will be included here.

Static Charges and Coulomb's Law

When electricity isn't moving, it's called "static electricity". This is the kind of electricity that gives you a little shock when you touch metal after riding a plastic slide or makes your hair puff out when you rub a balloon. The way to conceptualize static electricity is a world full of point-like "charges" which have a value (either positive or negative). In reality, these charges are actually the electrons in atoms, but due to a historical mishap a "positive" charge is actually a lack of electrons while a "negative" charge is a surplus of them. The way these charges interact is where things get interesting. In the simplest case, imagine we have two charges sitting near each other. How will they end up moving relative to each other? There is a simple rule for this:

• Like charges repel, and

• Opposite charges attract

So, if we have a positive charge and a negative one, they will be attracted together, while two positive charges or two negative charges will move away from each other. As we add more charges into the model, we end up with more and more attractions and repulsions which add together to create the overall effect on each particle. Now, the issue with the above rule is that it leaves out a critical detail: how much will they get attracted or repelled? This factor is accounted for in Coulomb's Law, which states that

F = k(q₁ • q₂) / r²

where F is the force (in Newtons) between charges q₁ and q₂, r is the distance between them and k is the Coulomb constant (around 8.99 • 10⁹ kg • m³ • s⁻² • C⁻²). The charges qᵢ are measured in Coulombs, a unit of charge denoted "C". This may look familiar — indeed, it is algebraically homogenous to Newton's law of universal gravitation (differing only by a constant factor). This is because the attraction between charges follows the inverse square law, just like gravity, and this form represents such a relationship. With this information, it is possible to determine how charges will move based on the charges around them.

Circuits — Series, Parallel, Voltage and Current

A circuit is simply a conductive loop. Conductive meaning that electrons are able to move along this path, and loop meaning connected to itself in some way (thus a figure-8 or other more complicated multi-loop paths are still circuits). To begin, let's imagine a very simple circuit: a circular loop of conductive wire. This circuit is not very interesting: all the electrons in the wire are simply sitting still. In order to make it more interesting, we have to add two elements: a voltage source and a current source.

Current is the amount of "flow" in a circuit — this is how "fast" the electrons are moving through the wire. If we imagine the circuit is a water pipe instead of a wire, this would be how fast the water is flowing. Voltage, on the other hand, is the "amount" of electrons moving in the circuit. In the water analogy, this is like the pressure of the water (higher pressure means more water in the same amount of space moving down the pipe). Luckily, we do not have to add these things independently into a circuit: both can be added at once by using a standard battery, which provides a fixed voltage (such as 1.5V for a AA or 9V for a 9V battery) and some amount of current. (The exact workings of how it does so is beyond the scope of this discussion, but in essence it is a chemical reaction which moves the electrons. This chemical reaction only lasts for a certain amount of time, which is why batteries don't run forever.) You might expect it to provide a fixed current, but this is not the case. Returning to the water analogy, a battery is like a pump. Obviously, if you connect tiny pipes to the pump and then ask the pump to move a huge water pressure (large voltage) through that pipe, it will not be able to move the water very fast (low current)! Conversely, if you ask it to pump that same water pressure (same battery voltage) through a very large pipe, it will do it with ease and result in a fast flow (large current). Here we see an interesting relationship: with a fixed voltage, current through a circuit will vary depending on the resistance throughout the rest of the circuit. This is more generally formulated in the form of Ohm's Law, one of the fundamental principles of circuit analysis:

V = IR

where V = voltage measured in Volts, I = current measured in Amperes (recall that Coulombs is C, thus current cannot be letter C) and R = resistance measured in Ohms. Looking at our previous example, if we hold V fixed we see that a larger R results in a lower I while a smaller R results in a higher I. Therefore, Ohm's law holds up intuitively, and it can indeed be verified with a few simple experiments.

Now, what provides the resistance? Well, in the case of our loop, there is none. So, if we add a battery into that loop (essentially, connecting the positive battery terminal to the negative one with nothing in between) our equation from Ohm's law becomes V = I • 0. Assuming the battery is not dead (0V), this would require the current to be infinitely high! Of course, in reality a wire cannot have zero resistance since the metal itself provides a small resisting force, but this resistance is so negligible that the current is still astronomically high. It is this exact phenomenon that is known as a short circuit — if you try this yourself, you will see that the battery quickly runs out of power when such an arrangement is held for any extended amount of time. This is because the battery is having to provide huge amounts of current which will soon use up all its energy. (Think of the energy used by a small pump attempting to move water at the speed of sound.) One important point is brought up by this thought experiment however: it is clear that there is some measure of the battery's total "energy" which should take into account both voltage and current. This measure is called Power, and it is measured in Watts (W) which is equivalent to Volts • Amperes (V • I).

If we don't want to continue pouring our money into wasted batteries, we need to add resistance to this circuit. Luckily, there are electronic components designed exactly for this purpose: resistors! Each resistor has a given resistance value, in Ohms, so by selecting a resistor and instead placing this resistor into the loop with the battery it is simple to see that the current will be much lower. (For instance, a 1 kilo-ohm resistor placed in a loop with a 1.5V battery will result in a current of 0.0015 Amperes, or 1.5 milliamps — a very reasonable current.) Of course, resistors by themselves don't actually do anything useful, so we can also add a light bulb or an LED (a semiconductor which generates light) into the circuit as well. Light bulbs have moderate resistance by themselves, so it would likely be safe to simply put the lightbulb alone with the battery (without a resistor). But, LEDs have almost no resistance inside, so when using an LED we must also put a resistor into the circuit. The question is... now that we have three components, how do we arrange them? There are two meaningful ways to arrange these three components: putting the LED and resistor in series or in parallel. A depiction of each is shown below (green indicates a positive voltage, and the lines are wires).

To decide which to use, we need to know how current and voltage are affected by these different layouts. (Obviously, we want the current through the LED to be the same as that through the resistor — otherwise, we end up with the same V = I • 0 issue, since the LED has effectively no resistance.) We need a new set of laws to define this, but first let's tackle this problem intuitively. We'll start with series.

Let's use the water model to imagine the series circuit. Water from the pump goes first to the LED, a nice wide pipe with little resistance, and then to the resistor, narrowing down to a thin pipe with high resistance, before returning back to the pump. What will be the current through each component? Well, even though the LED's pipe is wide, the water can't go fast since immediately after it must go slowly through the resistor's pipe. It's like having to slow down on a wide road because the road narrows later on — even though you could normally go fast, you can't because of what's ahead. Therefore, we see that in a series circuit, current is the same through all components. Now, what about voltage? Well, let's defer to algebra for a moment: since we know the resistance of the components, and we now know the current through the entire series circuit, we see that V = IR means in a series circuit, voltage is proportional to the resistance of the component. (Think about it: we have a fixed I in all the components, so we can ignore that if we say "proportional", and that just leaves "voltage is proportional to resistance".) Returning to intuition, it is clear that in the thin high-resistance pipe the water must be "packed tighter" than in the wide low-resistance pipe — thus there is higher pressure in the high-resistance pipe meaning it has a higher voltage.

Now, let's look at parallel. This one is a little trickier — let's start with voltage. How does the pressure change at the split? Well, it doesn't: there's simply two pipes now, but still with the same pressure. (Consider this one for a moment, as it is a bit less immediate than the previous conclusions.) So, in a parallel circuit, voltage is the same through all pathways. Hmm... this sounds familiar! Recall that in a series circuit, this is exactly how current behaves for each component. Perhaps, then, current will split proportional to the resistance of each pathway, just as voltage did before for each component. To verify this, let us once again turn to Ohm's law. We now know the voltage through each pathway is the same, so we can see that I = V / R (a quick rearrangement of the original law) means that in a parallel circuit, current splits inversely proportionally to the resistance of each pathway. (Our V is constant, so we have "I is proportional to 1 / R".) Intuitively, this makes sense: if the water is faced with two possible paths, one with low resistance and one with high resistance, most of the water will go down the low-resistance route and only a small amount will travel down the high-resistance route.

So there we have it — four rules that define how voltage and current behave in series and parallel. These laws are a somewhat alternative formulation of Kirchoff's Laws (they have been presented this way in the interest of explicitness). All that is left is to resolve the original question: should we choose parallel or series to make sure the LED has the same amount of current flowing through it as the resistor? Well, looking at our laws, we see that current is the same through all components in a series circuit — thus we know to put the resistor and LED in series in order to make sure too much current does not go through the LED. (If we were to choose a parallel circuit, the current would split inversely proportionally to the resistance of the LED and the resistor; thus almost all the current would go through the LED, as its resistance is quite low, making the current astronomically high once again.) Through a little algebra, we can also solve for resistance instead of current/voltage given these laws to simplify our conclusion: in a series circuit, the total resistance is the sum of the resistances of each component, and in a parallel circuit, the inverse of the total resistance is the sum of the inverses of the resistances of each component. (The second one is slightly harder to understand in words — it means that 1/[total resistance] = 1/[resistance of component 1] + 1/[resistance of component 2] + ...)

Of course, these two arrangements are only the start of limitless other possible combinations when you begin adding more components. Complex circuits can be made from many compound constructions of both parallel and series circuits together in order to create complex behavior. However, these same rules always apply, which allows you to analyse any circuit to produce the desired effect. Likewise, there are many other electronic components which can provide different functions (such a switches, which can stop current, capacitors, which can store charge, and transistors, which act like controllable "valves"), but the rules of voltage, current and resistance always apply to any component. (Unless it is a superconductor — then the previous argument that all components have resistance is false.)

Programming

By creating very complex circuits (very cannot be stressed enough), it is possible to construct devices which can perform multiple functions. For instance, imagine we have created a much simpler device which can do one of two things: turn on a light or beep a buzzer. Let us also imagine that the way to select which function it performs is to set a switch on the device to one of two positions. Now, let's suppose we have a specific purpose for our general-purpose-blinky-beeper™ in mind (a purpose which is, of course, turning on a light or beeping a buzzer). How would we tell the device how to do this purpose? In this case, it is trivial: simply set the switch to the desired function. This act of "telling the device what to do" is called programming.

Of course, our imaginary device is not very useful. So let's make a new device which can do one of three things:

• Turn on the light

• Turn off the light

• Check if a button on it is pressed, and if it is, skip the next command

At this point, the device is getting a bit more complex. And as hinted by the third command, it's going to need a modification: being able to execute multiple of these "commands", one at a time. So how do we program it now? Well, we can write down a "to-do list" for the machine and (supposing we have some kind of input parser for it) give the list to the machine to have it run. This form of programming is much more similar to what we use in the modern day — though of course, we do not use physical to-do lists (a.k.a punchcards) anymore, but we instead transfer the data to the machine via digital inputs. Returning to our imaginary machine, let's examine a possible to-do list for it:

• Turn on the light

• Check if a button is pressed, and if it is, skip the next command

• Turn off the light

So, we have two possibilities here. Suppose we have the machine run the to-do list when we are not pushing the button. Then, it will a) turn on the light, b) check if the button is pressed — which it is not, so the next command is not skipped, and finally c) turn off the light. Thus, the machine will just quickly blink the light (or if the commands are run fast enough, you will not even see the light turn on in the first place). What if the button is held down? In that case, it will a) turn on the light, b) check if the button is pressed — which it is, so the next command is skipped and there is nothing left to do. So, if we do hold down the button then the light will remain on. This illustrates a primitive form of decision making: our machine is able to decide what to do based on whether the button is pressed. We've successfully made a button trigger a light. How exciting.

Reflection

This project was probably my favorite out of the whole year, and I was very proud once it was "completed". I write completed in quotes because I do plan to continue working on it — there are still a few issues in the software which have yet to be resolved, and I would like to add a few more features to it such as an auto-position-reset (instead of manually winding up the strings each time). This, I think, indicates that the project was a definite success, and even in this form the game was quite fun! To begin, I think my critical thinking was a big strong point in this project. The development process was rocky, to say the least, but I managed to figure out solutions every time the problems presented themselves. For example, just 3 days before the presentation I discovered that the gearmotors I was planning on using to move the lander did not have the torque to run slowly enough for the precision needed. This was at first a large blow, as I had essentially completed the entire hardware setup. However, I quickly disassembled the motor mounts, designed new ones to 3D print out the next day, modified the pulleys with a Dremel tool and scoured around for some alternative motors. I found two continuous rotation servos in an old robot, which I hastily gutted and rewrote the software to accommodate them. While this meant I had to scrap the infrared encoders I was planning on using (modifying them for the servos would be a process that I did not have the time for, given how soon the deadline was), the motors proved to be a fairly suitable replacement and they ended up working fine.

Additionally, I think my conscientious learning was a positive aspect to this project. There were just 3 weeks — from initial conception to presentation day — to get the project together, and given the complexity of this design I was rather worried that it would not be complete in time. However, I worked really hard on it and put in a lot of time outside class to make sure that I would get it done. (Indeed, it was finished the night before it was due; had I not planned a few days as extra time to refine, the aforementioned gearmotor problem would have sabotaged the success of the project fatally.) In the end, the proper schedule planning that was done helped make sure the project was as polished as it could be. Finally, I also believe that my communication was good in this project. While the physical presentation was very informal, I think this write-up (and the other documents it contains) is a good indicator of this project's communicational strength.

As usual, there were a few small details which could've been improved. Given this was a group project, I think the collaboration probably could've been improved a little. I ended up doing a lot of work on this project, which was by nature as we did not have much in-class time and someone had to work on it outside of class; however, I should've planned my work more carefully so that I would know exactly what needed to be done in class to maximize what little time we did have. (This also factors in to conscientious learning.) Also, I think my conscientious learning could've been improved in the workflow of the project. Because of the aforementioned in-school/out-of-school dichotomy, we ended up writing a lot of the software before the physical hardware was built. This is a horrible idea since software is best built incrementally, and likely this design style might have led to some of the issues that ended up plaguing the landing detection system later on.

Aside from the few small issues, I really think this project was a big success. I planned an ambitious design, worked hard to see it through, overcame the challenges and ended up with a great final project. I'm really happy with the way it turned out, and hope to have more projects like it in the future.