ME 333 Final Project Ideas
- A ball-on-beam balancer. A ball runs on two rails, and a motor controls the angle of the rails in gravity, allowing the ball to roll left or right. Goals would be to stably move the ball from one position to another, or to have it follow a trajectory. The ball's position on the rail is measured by using the (low) resistance of one rail as a potentiometer and the ball as a wiper to the other rail.
- As a more advanced version of the ball-on-beam balancer,
make a robot "hand" that rolls a ball from its "palm" to the back of its hand
using a single continuous motion, much like a human juggler can do. This trick is
sometimes called a "butterfly." See the figures below for a human version and a
simulated robot version. To sense the ball's position,
you can use a metal ball as the "wiper" on a potentiometer
formed by the small resistance along one of the rails the ball travels on.
You can also make it so a player can try to do it manually.
See also this webpage at the University of Rome "La Sapienza" for a better idea of how to build one.
- A shoot-the-moon game, similar to an ME 433 fall 2005 project, which uses similar sensing to the balancer mentioned above. By opening the two rails, the ball rolls under gravity, and you try to get it to roll as far as possible before it drops between the rails. This project would include modeling the dynamics and figuring out a good strategy based on the dynamics (investigate if this problem has been studied in popular physics journals), and should allow the user to choose which hole the ball should be dropped into. Perhaps a user can play too, using an analog joystick as input.
- A magnetic levitator. A metal ball hangs below an electromagnet and is levitated by the electromagnet. The ball's position is sensed using an LED and a phototransistor or position-sensitive detector. Note: this is a very challenging control problem!
- Control of an inverted pendulum.
- Missile command. A user rolls a ball into a playing area, its motion is detected by paired laser beams and photodiodes, and another ball is launched (rolls down a ramp in gravity) with the goal of intercepting the first ball. Make it handle bank shots. Are there other good solutions for getting the (x,y) position and velocity of a ball on a flat surface?
- Track the location of an object moving in water by sensing its disturbance to an electric field, move toward it.
- Use four current-amplified oscillators of different frequency (in the audible range) to create oscillating electric fields in four buckets of water. Inserting an electrode into a bucket, you can pick up the signal, amplify it, and play it over a speaker. By inserting it closer to the electric field source, you get a stronger (and therefore louder) signal. Make an instrument out of this whose controls are the depths of the rods. Your PC can automatically play a song, or a user can provide manual input to adjust the depth of the rods.
- Elevator control. You have a panel of buttons which are for up/down on each floor and a panel of numbered buttons for inside the elevator (all on the same external interface). Your scaled-down elevator moves up and down and opens/closes its doors according to a smart scheduling algorithm. Might have a light sensor to keep doors from closing on someone.
____________________________________________________________________________________________________
When the rider encounters an uphill grade, they have to apply more force (torque) to the pedals to keep the bike moving along. At some point, the force they must apply is great enough to make them change gears to make it easier to pedal. Likewise, if they are peddling effortlessly on a downhill, they change gears to maximize the mechanical advantage of the bike.
Sensors can be used to detect the amount of torque the rider applies. The real-time control system is programmed with thresholds for each gear controlled (for simplicity, the system will only manipulate the rear gears). A motor is used to pull/release the derailleur cable that determines which gear the chain follows.
Testing/demo can be done with the bike on a stand with variable resistance.
____________________________________________________________________________________________________
Mission Statement: To create a two part system inspired by the cartoon Tom and Jerry.
Motivation: This product may be sold as an amusing toy.
Design Overview: There will be two parts to this system: a mouse shaped part, Jerry, and a cat shaped part, Tom. Both will be mounted with sensors and motors. They will communicate with each other, and Tom will chase Jerry. Tom will not be equipped with any sensors other than the sensor to detect Jerry, so that he bumps into walls and other objects (hardware will be protected) as in the cartoon. Jerry may or may not have other sensors to prevent him from crashing into objects, but will have a sensor that detects Tom so he goes in a different direction. The two might also communicate so that Jerryís velocity is always greater than Tomís.
____________________________________________________________________________________________________
I would like to make a reed that can deflect variably when touched, depending on how damped the system is. The "reed" would be a relatively rigid spring with a weight on top of it that would be one meter tall and 5 mm thick. There would be three cables running through the center of the spring, such that motion in the xy plane could be completely controlled. The cables would be anchored 180 degrees out of phase from each other at the top and bottom of the spring. This would maximize the length of cable that would be needed to winch the top of the spring to its fully deflected position. From the dimensions proposed, the cable would need to be winched by 12.5 micrometers. This would be accomplished using a good electric motor and a gearbox. Once the top of the spring had moved to its position of maximum travel, the weight would take over. Its centroid would hang past the furthest point of anchorage at the bottom end of the spring, slowly bending the spring over itself. This could be slowed down or reversed by winching the appropriate cable, returning the spring to an upright position. The speed with which the spring righted itself would be a parameter of the system.
The deflection of the spring would be detected by a ring of three IR sensors slightly above the point of anchorage. Each sensor would be connected to its appropriate winch, and would control the speed of rotation according to a function like this:
where deflection is on the x-axis and winch speed is on the y-axis. There are three regimes in this control function: At the left, the positive values for the curve would keep the resting spring stationary. After this, an outside deflection would cause momentary positive feedback as the reed exaggerated the impulse. Finally, the reed would begin to reverse the positive feedback and right itself.
____________________________________________________________________________________________________
My idea is to build a ëSip and Puffí Fishing pole. This fishing pole design would be geared toward paraplegic people. It would be a fishing pole controlled entirely with a mouth piece that you blow into, sip from and bite on. To explain the idea I will step through the process of casting and reeling in the lure.
First the user sips from the mouth piece to set a distance to cast the lure to. This sipping activates a motor that pulls the fishing pole back to a point, where a spring that is attached to the fishing pole is now at a certain elongated state. Once the user is ready to cast, they will blow into the mouth guard to release the pole and let the spring fling it forward thereby casting the lure.
Once the lure has been casted, the user can ësipí on the mouth guard to slowly reel in the lure. How much or how hard they ësipí will determine the rate that the lure is reeled in. Also, if they get a bite they can bite down on the mouth guard to reel in the lure quickly. When the lure is reeled in all the way to its ëhomeí position, the user can begin the process of casting all over again.
____________________________________________________________________________________________________
In a soccer game scenario, the user flicks a ball to imitate a cross, and moves a robot forwards or sideways in an attempt to hit the ball into the net at the far end. This robot consists of 2 players, similar to a table soccer setup, and its attempt to hit the ball in will be defended by 3 defenders and a goalkeeper, which is controlled by the system.
The ball is "crossed" by means of a single tunnel that it slides through by means of a flick of the finger. There will be four different tunnels allowing the ball to land into the danger zone and offering options to the user to try to hit the ball in. Near the start of each individual tunnel will be sensors that detect which tunnel the user has picked, and thereby predicting where the ball will end up in the danger zone. This will cause a movement of the defenders in the vertical direction perpendicular to the goal. On the other hand, the goalkeeper moves in accordance to the horizontal movement of the user controlled robot. In summary, the defenders move to cover depth while the goalkeeper moves to cover breadth (of the goal). The players will move via ërailsí inbuilt into the pitch such that only two motors are required for the defending side (one for the goalkeeper and one for the defenders).
Two of the tunnels will be meant for the ball to be a cross for the attacking players, while the other two tunnels will allow the user to attempt a direct shot on goal. The sensors in the latter tunnels will then direct the goalkeeper to ignore the input from the attackersí movement and instead go towards the side where the ball should land at, in an effort to keep the ball out.
____________________________________________________________________________________________________
From an overall perspective, the remote pen is exactly what the title would imply. The user would draw a "pen" over a sensor array, while a motor system would move an actual pen over a sheet paper in concert with the userís movements.
The project would consist of two main components: the sensor array and the motor system. The sensor array could be made from any number of components: optical sensors, pressure switches, infrared sensors, and the like. As long as the array can accurately determine the movement of the user, thatís all that matters. The motor system would be roughly a rectangle, with similar dimensions to an 8 by 11 sheet of paper. One motor would control horizontal movement, another vertical, allowing for full range of motion.
Controlling the motors will be very difficult based on the premise that the user can move his hand over the sensor array much faster than you can expect the motors to move. There needs to be a great deal of feedback and control to make sure youíre tracking, and not overshooting or oscillating around, the userís movements. Jerky movements would probably have to be filtered out, since theyíre most likely undesired anyway.
____________________________________________________________________________________________________
I propose creating a mechanism that will propel a ball along a desired track at a set velocity in spite of significant and varying air resistance. The setup would involve a glass tube in which the motion of the ball will be confined. The ball would run on two linear rods that act as potentiometers in order to sense the ballís position. One end of the tube would be closed off (setting the initial position of the ball), and the other end would be attached to an air pump. A schematic of the device is shown below:
The velocity and direction of the air being expelled by the pump would be controlled by a motor within the pump, in addition to possible external modifications to help in direction control. The motor would be run using a program created using Matlab/Simulink, or possibly altered manually. The position and velocity of the ball would be determined by a magnet attached to a moveable cart. This cart would utilize the feedback from the potentiometers to determine how much the ball has slowed down from wind. A Simulink program would integrate this information in order to instruct the cart how fast to go to keep the ball moving at a constant speed while the air pump is changing wind speed.
____________________________________________________________________________________________________
A well known situation: You're watching TV and eating some crackers. When you've finished eating you should put the waste away, but there is a distance of about 5m (!) between you and the garbage can, which would mean that you have to move your legs! Don't!
The GARBAGE TERMINATOR will sort out this problem.
The Garbage Terminator consists of two parts: The RUBBISH COMPRESSOR (a) and the CANNON (b). The compressor is a simple cylindrical construction which makes a piece of rubbish smaller and more compact. It's a strong motor which applies pressure on the garbage piece by pulling a heavy metal disk onto it over a wire. It is feedback controlled by sensing the torque (->current).
The cannon is movable in two angle directions, controllable by changing the reference signal coming from a joystick. It contains two encoders, one for each angle. It's the (funny) job of the user to choose the right angle position to hit the garbage can. The throw mechanism could be realized by a spring and an additional motor, whose job is compressing the spring.
____________________________________________________________________________________________________
A "motion" sensing light that only shines where the person is present. The application of this idea is to provide light in only a portion of a room. This idea seems like a practical idea, because it will enable one person to work in different locations in a room, while another person goes to sleep and is not bugged by the bright light that would normally be present.
How it works: Arrays of LEDís are placed in one half of a room (or for smaller scale, a constructed square), and photodiodes will be placed along the other half, so as each LED emits towards the photodiodes. When the path from the LED and photodiode is broken, the processor will compute where the location of the break is and focus light at that location. The light will be focused by using a lightbulb surrounded by a swiveling sphere with a mirrored interior. This sphere will be cut into one half or three fourths of a sphere, so that the light emitted from the bulb will be reflected off of the mirrored sides and towards the destination. The swiveling mirror will be controlled by a motor and a processor.
____________________________________________________________________________________________________
I propose building a wheeled robot which emulates a living creature whose primary goal is to obtain enough energy to continue to survive, but instead of food, the robot obtains energy from solar power. Ideally, the robot would be completely solar powered. However, due to the high cost of a solar panel capable of powering the PC/104 and motors, the robot will be powered by pre-charged batteries, and a low-power solar panel will be used to charge a small battery that will slowly dissipate power through a resistor. As long as the solar panel can provide enough power to keep the battery charged, the robot will continue to "live." If the battery runs out of power, it will trigger a switch that causes the robot to shut down, thus simulating the robotís death.
The robot will have multiple photodetectors to detect the intensity of light coming from different directions, and it will move toward greater light intensity. If the light intensities from all directions are the same and the robot is receiving enough power from the solar panel, then the robot will remain still. If the light intensities from all directions are the same and the robot is not receiving enough power from the solar panel, it will move around randomly until a source of greater light intensity is found, and then it will move toward it. The robot will be wheeled and will include sensors for obstacle avoidance to prevent it from becoming stuck.
____________________________________________________________________________________________________
My idea is an auto-leveler, for use in a number of different applications. It could be used to keep your drink from tipping over in the bottom of a canoe, or used to keep dishes from spilling on waiters' trays as they try to maneuever through a room. As the canoe rocked or as the waiter moved his tray, the auto-leveler would compensate and keep the contents from spilling.
Essentially, the auto-leveler would involve several sensors positioned around the tray that detected a tilting motion and measured the angle from level. I am not yet entirely sure how this angle would be measured or what kind of sensors could be used. Pressure sensors might be able to measure the changing component of weight, which could then be used to calculate the angle. However measured, the angle would then be transmitted to any number or arrangement of small pneumatic cylinders that would then counteract the tilting motion.
____________________________________________________________________________________________________
My proposal for a project is for a projectile device. I would like to attach a nerf foam dart gun to a mechanism that would aim it at a target and fire. Optical sensors would put lasers or the like on the object in view, each laser being indepedently controled. The vectors of each laser would be sensed and the position of the object would then be calculated using the the two laser vectors. The computer would then calculate the postion and direction needed to fire the dart in order to hit the object. Servos would probably be the best way to move and am the laser as well as the dart gun. And then a larger motor to "cock" the nerf gun as necessary, depending on the workings of the nerf projectile launcher. Im not sure exactly how the light sensors would work, but if that is no problem, I see this project being rather feasible.
____________________________________________________________________________________________________
My idea has to do with the food industry. A conveyor belt carrying whatever item would have size and/or weighing devices. Depending upon the weight of the item, there could be a rerouting of the product to another conveyor or eliminating it completely from the process. For example, with chicken, depending on the chickenís weight they could be sent to different conveyors for different package sizes or whatever. For soda cans, also, if the weight is too low or too high (depending on some normal range) the can could be rerouted somewhere to take it out of the process so that a bad product doesnít get out to stores. For the class project instead of using a real conveyor and real items we could use small random items.
____________________________________________________________________________________________________
Currently in the field of prosthetic devices, there is no knee replacement that allows the user to walk up and down stairs easily, if at all. For a client with a transfemoral amputation, we want to design a new model that allows the user to activate a motor with a button. This motor will cause the knee to rotate until the user is standing straight or at an angle equal to 0 degrees. The motor will sense the intitial angle, rotate until the angle is 0, then stop. Each time the user climbs a step, he/she will have to push the button to activate the motor. Our motor will have to create about 411 N*m of torque for a 200 lb person.
____________________________________________________________________________________________________
Build a maze, and have a robot navigate it. The maze should be comprised of "blocks" of uniform size, each block large enough for the robot to turn around in. Controlling the robot will involve:
* Sensing where walls are and where the center of the maze "block" is
* Turning and moving from block to block
* Navigating the maze - i.e. locating the exit or finding some object and returning.
* Battling the Minotaur (optional).
____________________________________________________________________________________________________
I am thinking of doing a twist on last year's XY plotter. Essentially it would be a robotic arm holding a pen that could draw arbitrary shapes onto a flat surface. It would have two degrees of freedom by having an "upper arm" and a "lower arm", each actuated simply by motor rotation. Sensing would be required to make strokes because the computer would need to be aware of the arm's position in order to compute the right angular velocities for the motors to move the pen in the desired directions.
The arm would be similar in size to a human arm, configured such that at rest position (x=0, y=0) the upper and lower arm would be at right angles to each other. To ensure stability I would constrain motion of the pen to within a reasonable area, perhaps 8.5x11" for drawing onto ordinary paper. I would design the code to treat arm angles as some continuous functions of the pen position: this way there would be only one arm orientation solution for any given point on the plotting area, and the arm would never to "spike" angles for small output changes.
To begin with it may be easiest to define some simple curves like parabolas or sinusoids within the computer. If feasible I would be interested in taking greater advantage of the sensing by allowing physical input: a human user could set the machine to "record", and manually the arm along an arbitrary curve. The computer would take regular samples of the locations (or velocities, or both) to remember the curve taken, and then attempt to repeat the motion without human guidance, perhaps with a differently-colored pen. Another application would be allowing a human to make large strokes and have the machine recreate miniaturized strokes to facilitate the creation of high-precision hand drawings.
I imagine that the entire thing could be done with readily available resources such as sheet metal, motors, joints, wheels and belts, and the typical circuit elements.
____________________________________________________________________________________________________
My idea is mostly borrowed from the elevator suggestion that is posted on the website. The idea is for elevator control. Instead of a scaled down people elevator I propose something more like an automated dumb waiter which would be useful in a home or dorm with a loft or bunked beds. There would be a panel outside with buttons for up or down. Instead of a numbered control panel, the elevator would be equipped with light sensors and/or weight sensors. The light sensors would be a safety so the doors donít close on a hand or on food. They may also aid in where the elevator would stop. So when the elevator sensed a dark spot (signifying a new level) it would stop.
The weight sensors would be there to control the motion. Depending on the weight sensed inside, the elevator would continue moving in its current direction or go back to its original location. If this weight sensitive control is too complicated the idea of having the numbered buttons on the panel with the scheduling algorithm could be used for the control. There could also be a platform inside the elevator that extends out when the elevator doors open and retracts back either within a specified time interval if nothing has happened or after a weight has been added.
Honestly I have very little idea of what my group and I are capable of. I also am not really aware of how feasible or what equipment this project would need. Obviously I mentioned weight and light sensors. Presumably we would be able to build some sort of small elevator. We would need motors for the motion as well as either a shaft or pulleys to guide the motion of the elevator.
____________________________________________________________________________________________________
There are numerous devices for changing light bulbs on high ceilings that consist of a method of grabbing a bulb and thenrequiring the user to twist the bulb out. This is often tedious and sometimes doesnít work if the base were to unscrew instead of the bulb. I think we should create a device that when raised up to the bulb grabs the bulb AND the fixture, unscrews the bulb, and screws in another.
The device would have to accommodate a range of bulb and fixture sizes, know where and how hard to attach, when to start and stop turning the light bulb, and switch from removing to inserting a light bulb.
A relatively new technology allows completely blind people to ëseeí via a soundscape generated by images captured by a camera. The 2D image is converted into a whole pallet of sounds where the x-axis is represented by how left or right the sound is in stereo headphones, the y-axis by the pitch of the sound, and the brightness of each location by the volume of each pitch. More info and the existing software is available at http://www.seeingwithsound.com/. One of the difficulties with the technology is training oneself to more accurately discern where a sound is from left to right and from low to high pitch. My proposal is a game that emits a moving target in this kind of soundscape and the user is given a gun and needs to shoot at where the object should be. Sensors in the gun would have to communicate with the orientation of the userís ears to know that the user has fired in the correct direction. If sound production doesnít count as actuation, then perhaps the gun can have a device the simulates recoil.
____________________________________________________________________________________________________
The mechatronics thermostat will take in a user-defined temperature. Then it will use a temperature sensor to read the temperature in the room. If the actual temperature is lower than the desired temperature, a space heater will be turned on, and the window will remain closed or will shut depending on it's current position. If the actual temperature is greater than the desired temperature, a fan will turn on and the window will open or remain open depending on it's current position. There will be a temperature range of about ±2ƒ where nothing will happen due to the fact that we do not want the machine to be constant running.
____________________________________________________________________________________________________
My idea for the final project is to create a XY-plotter along with a Z-axis movable electromagnet that would relocate metal ball bearings from a storing location into a predetermined pattern. This simple pattern could be input through an image file, or something similar. Different color bearings could be used to create multicolored images (up to 6 or 7 colors).
____________________________________________________________________________________________________
Such a device would operate continuously, motors would spin a central wheel, bringing the gyroscope to upright position. As the disc slowed its balance would waver, sensors would detect this and increase output to the motor to cause faster rotation, thus righting the gyroscope. Device should respond to human attempts to upset it's balance (one would move the mounting underneath it as a challenge).
A system of actuators would be mounted onto an existing bass guitar. Mechanical fingers would pluck the strings and depress the strings at the frets as necessary to create music. "Songs" would be programmed in, somehow I suspect this is way beyond our capacity.
Such a device would harness mechanical energy from a hydrualic flow (like falling or running water) and orient its position in the flow to maximize efficiency from a stream of varying strength. Could be based on regular waterwheel turbine, or more advanced turbines (gorlov helical or horizontal rotary bladed), focus would be on position control within stream, performance measured in watts of energy harnessed per kg water passed through turbine.
This device would follow the sun across the sky in order to maximize solar input throughout the day. Consists of solar panel mounted on rotating housing with sensors to determine position of sun and position of panel. Ideally would harness enough power to run itself.
____________________________________________________________________________________________________
Essentially, the system would take an assortment of socks as its input and output pairs of neatly rolled up socks. The mechanism would analyze socks individually based on differentiable characteristics: length, color, thickness, texture, etc. The machine would then match up socks with similar characteristics, then proceed to fold them. My idea is to have the machine pull one sock at a time, sense and record the sock's characteristics, and then place the sock in a "tray". If the machine detects a similar sock, that sock is placed in the same tray, and both are taken to some folding mechanism (perhaps built into the tray). The machine would have say...10 such "trays" (and hence capable of sorting 20 individual socks at a time).
____________________________________________________________________________________________________
My idea is to create a robot that imitates the movment of an inch worm. It would start out in an upside down U shape and then one end would leave the ground and stretch out. then reatach to the ground and then the other end wwould follow it. To attach to the ground it would need a suction system as well as a sensing system so it knows when it hits the ground. also a stepper motor or two would be needed to move the inch worm. I was thinking that a motor that reeled and unreeled a wire would be how the contracting and releasing of the inch worm would work. the robot would have to be made out of a segmented body or flexible material to allow this type of movement. I was also thinking that if possible and time permitting we could design the inch worm to follow a line. This would involve a lot more sensing and a much more complex design of the body so i am not sure it is a possiblity in the 5 weeks we have.
____________________________________________________________________________________________________
One idea is to build a maze and see if we can create a mouse that is capable of getting through it. The "mouse" would obviously be very simple: a motor on wheels, essentially. It would need to be able to make 90 degree turns in tight areas, so perhaps the best design would be to have four wheels that can rotate 360 degrees. A square body with sensors on all sides might be practical in this case, so that any side of the mouse could be considered the front. If the wheels rotate 90 degrees to the right, the body of the mouse doesnít move but now the right sensors are considered to be the front sensors and the front and back are the side sensors. The wheels would only need to rotate in one direction with this design.
The mouse would somehow need to be able to sense where the walls were. One possible way to do this might be to give the mouse a laser beam to shoot straight out in front of it and also a photodiode. If the walls are reflective (mirrors), it will know how close it is to anything in front of it based on the time it takes for the reflected light to be detected. This could be quite extensive, though. Another possibility would be to just let the mouse run into the wall and with force sensors, then, detect that there is an obstacle in front of it. I donít know if there are other types of sensors that might be better for this type of sensing..?
Finally, an extensive control algorithm would be needed to provide the mouse with the intelligence it needs to get through the maze. The algorithm must ensure that the mouse does not ever backtrack and go down paths it has already tried, so there should be some strategy behind which directions the mouse chooses to turn. There also needs to be control algorithms in place that take into account the information collected from the sensors on the mouse and move it accordingly.
____________________________________________________________________________________________________
My idea for a mechatronics project is a basic force-feedback steering wheel system. This will consist of a display that either has a straight path or a left or right turning path with a small car, a pedal, and a steering wheel. The game will be made so that when the car is turning, the further the wheel is turned and the faster the car is going, the harder it is to turn the wheel. The pedal is just a pressure sensor, and this data combined with an angle encoder attached to the wheel is used to calculate the torque that should be applied to the steering wheel through a motor (proportional to the speed of the car and angle of the steering wheel). This gives resistance to the user when trying to turn the wheel. The feedback would have to be at a pretty fast rate as the motor is not meant to turn the wheel, only to provide resistance when the user is turning the wheel.
____________________________________________________________________________________________________
The robot consists of a small car with a camera attached. The camera observes its surroundings and transmits a wireless signal to at the host computer.
When the computer recognizes a bright laser or infrared light (activated by the user) with the camera, the robot moves to the target and follows the light. Additionally, a touch sensor and two arms can be attached to the robot. This would allow it top grab the target and bring it back to the starting position.
____________________________________________________________________________________________________
The project I propose is to balance a marble on curved track. The track will have a convex curve so that there will be no steady state for the marble. To balance the marble, the angle of the track will be measured by an encoder and controlled by a motor at a pivot point below the track. I'm not knowledgeable about sensors, but I would think infrared sensors could be placed to sense across the track and sense the location of the marble. Then the computer would calculate the velocity of the marble, and adjust the track to counteract the movement.
____________________________________________________________________________________________________
My project idea is based on the childrenís board game Mouse Trap. Basically, I would attempt to build an electrical/mechanical mouse trap. I envision the project using four fixed lasers aiming at four fixed light sensors. These four lasers will be arranged so that their light creates a small (about 3" by 3") square. Inside of the square would be "bait" for the mouse, such as a piece of cheese or peanut butter, etc. When the light sensors sense a change or interruption in the laser light source (an object breaking the connection between laser and light sensor), it will trigger the release of a cage.
The cage would be made of steel and there is much liberty as to how it is constructed or what it may look like. The important part mechanically will be the way in which the cage is dropped. I envision the cage being attached to a lever arm, being to rotate in one direction at the end. When in the upright position it would not fall down due to a "catch" that would hold the arm up. When the sensor detects a break in the laser it would send a signal that would pull back this "catch" (operating somewhat like a trigger). The cage would then fall over the area encompassed by the four lasers. One problem I can see with this project that I would have to work around would be making the trap fall quickly enough to catch a mouse while at the same time not having it crash violently into the ground and breaking.
____________________________________________________________________________________________________
The airplane game is basically a joystick controlled physical "airplane" that is attached to a small metal wire that keeps it "in the air" and you control its x, y, and z coordinates while traveling in a "world". The "world" is really just a rotating barrel with obstacles that can be knocked down if hit. If they are knocked down, then a sensor senses it and counts it as damage against the plane. As the world rotates, the knocked down obstacles are reset by gravity. We could also add some enemy plane AI. That come in from behind or in front. Or we could make it a race, and have the other planes controlled by computational programs that dodge the objects by themselves using laser diode/ phototransistor pairs. If they fall behind, they just disappear from the bottom of the screen, if they pull ahead, they disappear from the front of the screen.
____________________________________________________________________________________________________
An "artificially intelligent" (or at least, so to speak) computer player for the game the operation. If you're not familiar, see this link.
Players alternate pulling bones out without hitting sensors. There will be a crane for the computer player, obviously. With motor actuation and control, the crane can be moved over a bone and then lowered to grab it. Sensors can detect what bones remain so the computer doesn't choose one already taken by either the player or itself. Grab the easiest bones first, or maybe the most points first.
It might have to bend the game a bit and make the bones magnetic so it is easier to grab them. The important part is that you can offer various tolerance levels, either in the distance from the bone to the body (meaning you lose) or even offer different speeds at which the crane operates with worse and worse motor control so that perhaps it has some overshoot or oscillation on really easy settings, and on hard settings it takes its time.