meow the mechanic

Fashion Show

2010-06-26T07:35:00.000-07:00

A fashion show where designers and models are all students.

Chi Nin Nunery Second Visit

2010-06-20T06:46:00.000-07:00

It's hot like hell when shooting under the sun. However, its a great place for photography.

Central Historic Landmark

2010-06-19T07:23:00.000-07:00

Earliest historical buildings are all in the Central to Western district in Hong Kong Island. Although many had been torn down in the last century, there are still a number of them standing quietly there decades after decades.

Mosque

2010-05-23T07:43:00.000-07:00

A guided tour to the Mosque at TST, very interesting.

Louis Ku's Collection

2010-04-11T07:44:00.000-07:00

Mostly Star Wars figure, with some others such as Spiderman, Terminator and Alien.

A trip to Tsoi Yuen Village

2010-03-20T07:38:00.000-07:00

Not much photo to shoot, anyway, they told us a story about the tear down of a village.

Yuen Long Park

2010-02-20T06:19:00.000-08:00

Almost stepped on it, kind of frog, but not exactly, its not green colored.

Did you ever saw a turtle stretch its legs so straight??

Dance Day

2010-02-06T05:51:00.000-08:00

To get a good location for photo taking, you have to be there early, otherwise there will be a lot of heads in the foreground

Tai O

2010-01-16T05:35:00.000-08:00

At this time of the year, take pictures at Tai O between 4:00p to 5:30p will have very nice sun light effects.

Ngong Ping

2010-01-16T05:26:00.000-08:00

For photo taking purpose, buy a single ride cable car standard ticket is the most cost effective, take pictures from the cable car, then the market place and the budda, finally, take bus to leave or go other places in Lantau Island.

Golden Coast Beach

2010-01-09T06:35:00.000-08:00

It was still cloudy in the morning, but all of a sudden, the sky turns blue in the afternoon, and even with the bonus of fluffy clouds.

NXT Claw

2009-12-31T05:57:00.001-08:00

It has been over a year I have not created any stuff out of the Mindstorm NXT, now I am back and need something to warm up. A quick and simple project is something I wanted, here it is, the NXT Claw. Put your hand on the swinging platform, it may attack you, or just make some fake attack. The rule is like this, if you retreat successfully on a real attack, you win (and hear some win tones). However, if you retreat but it was only a fake attack, you lose (then comes the lose tone).

The mechanism is simple, motor driving the claw, going up and down. The player put his hand on the swinging platform and press it lighting until it is horizontal (there is a hard stop to help you feel when it is horizontal). Finally a light sensor which is either blocked by the platform when horizontal (hands on), or unblocked when the platform drops (hands off). The platform is purposely made to be unbalanced, so it drops itself by gravity when hands off.

The program is written using RobotC, the logic is like this:

if light sensor sensed it is hands on, a random number is generated, whether the claw should stand still, attack or just make a fake attack will depend on that random number
if fake attack but hands off, play lose tone, otherwise play win tone
if real attack but hands on, play lose tone, otherwise play win tone

You may further improve the program by counting the scores and see who did better. Also, the motor is programmed to move around 90 deg so as to raise up and come down, there is no reset function and I wonder if after many many rounds the motor still keep running at those angles. However, I am not going to add this as similar things had done before, and don't want to make it too complicated (I'm not making an end product for sale, so less worry about these kind of details).

As for the sound files, I feel ok with the "lose" sound, which is available in the NXT brick. However, for the "win" sound, I like something else, so I downloaded a program called wav2rso to convert wav file to rso format. Do a Google search for wav2rso, its easy to find it. The one I converted can be downloaded here.

You may download the program from here and add more logic to it, such as calculating score, alter the probablity of fake or real attack...etc. Also, don't worry if you lose as it won't hurt. I also provide the LDD file for reference.

Lei Yu Mun Resort

2009-12-25T06:45:00.000-08:00

It was once used by soldiers, now a public resort, quite a nice place to take pictures when weather is fine. However, the food at its canteen is less than satisfactory, and residents are not allowed to walk around after 10:00pm.

Models

2009-12-20T01:38:00.000-08:00

They practice on make up, we practice on portrait photography, both learned a lot.

Central Night Scene

2009-12-19T07:15:00.000-08:00

Longer exposure, more cars, the better.

Clock Tower and HK Cultural Centre

2009-12-12T03:19:00.000-08:00

The two building looks nice when appear together in a photo.

Former Marine Police Headquarter

2009-12-12T03:15:00.000-08:00

Now is a restaurant, beautifully lighted in the evening.

Christmas Lightings

2009-12-12T02:46:00.000-08:00

Traditional Christmas lighting seems more camera friendly, those new ones made using LEDs and changing color from time to time results in unpredictable color.

Ma Wan Park

2009-12-06T06:21:00.000-08:00

As seen in the Noah's Ark.

Eagle

2009-12-05T07:05:00.000-08:00

Inside Kadoorie farm, a lonely eagle standing there thinking about something.

Kadoorie Farm

2009-12-05T05:05:00.000-08:00

Climb uphill continuously for an hour, then look downwards, you will see this.

Christmas icon of 2009

2009-11-28T05:58:00.000-08:00

Combine the design of Hello Kitty, bee, Santa Claus and fairy then you got this.

The chairman's seat

2009-11-28T05:53:00.000-08:00

All you have to do is listen to B.S. and tell them to shut up.

Legislative Council, corridor

2009-11-28T05:50:00.000-08:00

While queue up for the tour.

Bethany Chapel, side view

2009-11-28T05:45:00.000-08:00

10mm perspective, very nice!

Bethany Chapel

2009-11-28T05:38:00.000-08:00

Exceptional good weather that afternoon.

Moon

2009-10-30T06:26:00.000-07:00

That's the size of a moon shot with 250mm, 1.6x crop body, 100% crop from 3888x2592 pixel photo.

Sunset, almost

2009-09-06T06:24:00.000-07:00

After a days work, the sun is tired, need to hide himself for a few hours.

Shallow river, street lights

2009-09-06T06:16:00.000-07:00

Two birds, looking for fish in shallow water. Not much people walking nearby, it seems the street lights were made for them.

Sunset, river and wetland

2009-08-24T06:13:00.000-07:00

A day with haze, still look good.

Sunset and river

2009-08-21T05:57:00.000-07:00

On an extreme fine day, may only once or twice a year...

Flower, pair

2009-08-17T03:19:00.000-07:00

White flower, dark background.

Golden sunset

2009-07-13T06:32:00.000-07:00

Happens once in a year.

Clouds painted on the sky

2009-05-17T06:27:00.000-07:00

Who painted it?

Mr. Lizard

2009-05-09T06:32:00.000-07:00

Its tail is longer than its body.

Trees, buildings and perspective

2009-05-09T06:27:00.000-07:00

Trees vs buildings, road vs grass, nature vs man made...

My Relatives

2009-01-18T06:05:00.000-08:00

I met my relatives in a Sunday afternoon.

Branches on blue sky

2009-01-03T05:54:00.000-08:00

Nice blue sky, with a network of branches on top of it.

Red and green

2008-10-07T05:29:00.000-07:00

One flower or two flowers? pedals or leaves? I'm not sure.

Beautiful skin

2008-10-07T05:26:00.000-07:00

Very colorful, attractive, beautiful, ... << you put some more adjectives here

Prehistoric animal?

2008-10-07T05:21:00.000-07:00

Oh, this thing looks like decendents of prehistoric dinosaur...

Flowers, upside down

2008-05-24T02:56:00.000-07:00

Its really upside down, I had not rotated my camera.

Red flowers, upside down

2008-05-13T03:00:00.000-07:00

Another upside down flower I found in wetland park.

Monkey at Shing Mun Reservoir

2008-02-10T06:01:00.000-08:00

Monkeys waiting to rob people passing by.

Biped

2008-01-26T06:48:00.000-08:00

It is interesting to see a robot walking on its legs, I have seen some humanoid toy robots with 8 motors (or degree of freedom) in their legs so as to walk naturally. With only one set of NXT, all you have is 3 motors, a biped built using 3 motors is sure to have very "robotic" movement. There are two basic things you need in building a biped, first is shifting the weight from one leg to another, second is the walking motion. In this project, most of the time was spent on the mechanism design, I had gone through five different designs before finally came up with this working one.

A motor is used on each leg (let's call them B-C motor), they were mounted directly on their 'feet' and serve to shift the body weight of the robot, which mainly come from the weight of the NXT brick. The third motor (A motor) controls the walking motion, it strides the leg forward when it is raised by the weight shifting motion as mentioned above.

To shift weight to the left, the left motor turns so that the whole body incline to the left side, until the stopper hits the foot. While the left motor starts to turn, the right motor rotates in such a way that it helps to raise the right leg.

After the right leg is fully raised, the A motor turns and the linkage mechanism drive the left/right legs forward and backward to create a walking motion.

The biped can walk ahead as well as turning left or right. See how the legs walk and turn here when the ultrasonic sensor detects obstacle in front.

Due to the way the legs are connected, there is a bit of variation in the width between the legs when it walks. As a result, some clearance is required in the cross axle.

As usual, I prefer to remove the NXT brick easily for battery changing, so I just mount it by two cross axle beam. Note that bushes are added to avoid the brick bottom rub against the A motor.

Since it took quite a long time to came up with this design, I have documented it using Lego Digital Designer. The LDD model file can be donwloaded here, you can rebuild the exact thing or improve it by studing the construction. In the model, I did not mount the brick and A motor into the biped assembly as it becomes difficult to study them when assembled.

Note that the legs have to reset to a correct initial position before walking, after the program runs, press the left right key until the "4x2 angle beam" inclines forward. When it is ok, press the orange button again to start walking. If this initialization is not done correctly, the robot may walk backwards

I have made two versions of program for this biped, one is for autonomous walking ahead, with the ultrasonic sensor detect obstacle and turn right. The program is written in RobotC and can be downloaded here.

The other version is by using the blue tooth capabilities of Roborealm and RobotC, you can remote control the walking of the biped through a joystick on your PC. Download Roborealm program and RobotC bluetooth remote control program.

Ballad Dance

2007-11-24T05:46:00.000-08:00

She seems like weightless when dancing, he can spin her around so quickly and easily.

Robot Arm

2007-07-18T06:09:00.000-07:00

I am always fascinated with those industrial robot arms which can run smoothly and accurately through curves in 3D space, here I build my own robot arm which can run point to point, learn its movement through teaching it (moving it manually), and it will playback the movement (approximately only).

The design of this robot arm is quite simple, first get a motor (let's call it motor A) to control the panning of the robot arm. As shown in the picture here, motor A is fixed to the base frame and installed with a 24 teeth gear. It drives the large gear plate which holds another motor (call it motor B), with the third motor (motor C) mounted on motor B's shaft. With this construction, we have an arm like structure with can rotate and pitch/bend. As all I have is the basic three motors in the standard retail kit, I don't have anything to mount in the end effector, just put in a few pcs of angled beams.

When the motor starts up, motor B will reset to its upright position, the reset is done by the angled beam structure mounted on top of motor B. When motor B resets, motor C will move upright and touch the stucture, as the structure stops motor B from rotating further, the motor stall can be used as signal to stop the motor in the RobotC program. Actually, I referenced the motor stall sample program which comes with the RobotC package. It checks to see that if the difference between the current encoder reading and the previous encoder reading is the same, if so, stalling occurs and start to count the number of stalls. You can then define a certain stall number to determine stopping the motor.

After motor B has reset, motor C starts to reset also, it also hit against the same structure of angled beams and stops the arm in a horizontal position. As for the stall number, you may have to experiment a bit as we reset it slowly by sending less power, and the less powered motor could be more easily get stalled which creates a false signal.

Similarly, motor A resets until motor B hits against the stop, which is mounted near the end of motor A itself.

State of the robot arm when all motors resets. Motor A will then rotate back a certain angle and ready to get into the learning mode. All motor is still holding in its position with power. When the orange button is pressed, power to all motors are removed, and the robot arm enters the "learning" mode. The program is actually recording the encoder reading of the motors, this recording is make whenever there are changes in encoder readings from any one of the three motors. The recording continues until the left button is pressed, then the robot arm resets and starts to playback the motion.

Before playback the motion, the data recorded go through a processing which identifies the end points of motion. Non-end point data are throw away as it is difficult to follow every point of the data to playback, that's why the playback is not exactly follows the speed between the extreme positions. The end points are then used as encoder targets to command the motor's rotation.

Video showing the start>reset>record>reset>playback (and reset>playback again) sequence. The program will infinitely loop the reset and playback until the bottom grey button is pressed to abort it.

Testing the accuracy of motion, again, due to motor gear backlash, there are quite a considerable amount of error, the error from motor A has the most impact on the positional accuracy as the length of the robot arm magnifies it. If I have compass and tilt sensor, I may reduce the error better than using only encoder reading to assume the robot arm had reached a certain position. There are other things for further studies, they are:

- how the motor can follow a series of recorded points and run through them smoothly, which is speed and position control of motor through all the recorded points instead of just the extreme points

- reduce the mechanical backlash and improve the robustness of the structure

I posted my RobotC program here for reference, it is not optimized yet and I believe a number of similar functions can be generalized as they are more or less the same for the three motors.

Mybot

2007-03-26T05:36:00.000-07:00

The Tribot provides lots of fun, thanks to its mobility and the interesting gripper. However, I would like to have a compact vehicle robot that can patrol around, and also easy to replace batteries (due to NXT's apetite on battery power, need to replace my rechargeable batteries often). Also, I would like to move the motor-A to somewhere I can use it instead of beneath the NXT brick. As a result, I assembled the Mybot. In this project, I also used RobotC to program the NXT, the NXT brick is now running the RobotC firmware instead of the original one for NXT-G.

So, here is the design I have made, I put the motor B and motor C on the bottom, motor A is now placed in front of the NXT brick. A bumper is also made so that when it bumps into something, it will turn back and run again. As seen on the picture, the ultrasonic sensor is mounted on the motor A shaft, it could be mounted on some fixed location but I just mount it there to experiment various angle relative to Mybot.

To test out RobotC, I have written a program to let it detect the distance (using the ultrasonic sensor) from a wall, and keep running at a certain range from the wall. Ideally, it should follow the wall no matter it is flat or turn around corners.

Here is another viewpoint, showing the front side of the Mybot. The ultrasonic sensor is set at 45 or 60 deg from the Z axis (lets define the Z axis as the one running from front to rear of the Mybot).

The program logic runs like this: when the programs starts, it detects the current distance of the Mybot from an obstacle, compute the upper and lower limit of distance (as it will never run on an exact distance, therefore, a range is required). To run straight ahead, the distance measured should be within the upper and lower range. If the distance measured is greater than the upper range, it will turn right so as to reduce it, otherwise it turn left to increase it.

Top view of the Mybot

As seen in this picture, the NXT brick is just placed on top of the frame, nothing is used to fasten it to the body, so I don't have to dismount anything when changing batteries.

Bottom view of the Mybot, note that I found an empty space where I can install the light sensor, it is not necessary in this project, but installing here make the design looks compact. The touch sensor is installed as shown, whenever the bumper lever bumps into something, it presses the touch sensor to signal the run backward routine.

Back side of the Mybot, note that I used the same caster design as seen in the Tribot assembly.

Here is the video of how Mybot follows the 'wall', my apartment is small and full of stuff, therefore, impossible to find a wall to follow, so I put two big plastic boxes for it to follow. The RobotC program can be downloaded here. The current Mybot has the following limitations:

- when put it too close to a wall, it will have problems when turning corners, by experiment, it works fine when the reading is 20-30cm

- sometimes it does not follow corners properly, may due to no feedback from ultrasonic sensor (unknown cause), or the turning cause its distance far from the wall, which keeps it turn right continuously, this could result in bumping into a wall (as shown in the video), in this case, the bumper can come to rescue

- there is an unknown bug which cause the 'bump and go backward' action performed twice when a bump occurs, it is yet to fix.

NXT Tribot

2007-01-22T06:16:00.000-08:00

The tribot was build exactly (99.99%, tell you what's not exact later) follows the Lego documentation, I used it to experiment how things are detected using the ultrasonic sensor. While thinking for the project, I read Krystian Majewski article about a NXT radar, it inspire me to make the tribot scan its surroundings, look for an object and then go and grab it. Just want to share it here for all the fun I have in making it.

After I build the tribot, I have an idea of scanning the surroundings with the ultrasonic sensor and look for something, say, within 10 inch of distance. I first tried to do this by incrementing the B/C motors a degree a time and scan the distance, however, it seems that the data read back fluctuates a lot and it is difficult to determine when it "sees" an object. To get better understanding of the problem, I build a simple program in NXT-G to control the B/C motors to rotate slowly and scan the readings, save them to a file in the NXT and uploaded to my PC, plot them using Excel and the graph on the right is what I found.

Then I tried another approach, the B/C motors were controlled to rotate anitclockwise slowly, at the same time, scans the surroundings for objects with distance less than 10 inch. This is not accurate and it only serves to get a rough starting point, then, I use an iteration method to converge the result. The purpose is to find the point where the ultrasonic cone first intersects the object edge, then we will rotate the robot a known amount so that the robot faces the center of the object and grab it.

When the first point was found, the robot is stopped immediately. Depends on the speed of motors and other unknown reasons, the first point could be close to the edge of the object or already get inside the object a certain amount.

The robot is then rotated back (clockwise) a certain angle, check the ultrasonic again to see if it intersect the object. If intersection occurs, it means that the robot overrruns too much before detecting the object the first time. As a result, our next rotation should be also clockwise with the same angle as we have to find a point where the ultrasonic cone does not intersect the object.

If no intersection occurs, it means that the object edge is somewhere between this point and the point we first found it in our previous step. We shall then rotate anti-clockwise for half the amount and check again for intersection.

We can summarize the method as below:

1) Rotate anticlockwise to find the first point

2) Rotate clockwise for a certain degree and scan, if intersect, further rotate the same degree clockwise and scan again, until we found the point with no intersection.

3) If step 2 does not intersect, rotate it anti-clockwise for half the angle in step 2 and check ultrasonic distance. If no intersection, repeat again until we found intersection point. If intersection occurs, rotate it clockwise for half the angle and check for ultrasonic intersection.

4) Steps 2 and 3 are repeated according to the result of ultrasonic intersection with the object. It will eventually converge to a point where the robot motors no longer rotate.

The video below shows how it happens.

I take a small bottle as object for grabbing, the blue ball which comes with the nxt kit is too short, and ultrasonic sensor did not detect the ball well. I put a few coins inside the bottle to make it heavier, so that it will not fall down when the tribot collide with it. The bottle I use is a bit too tall for the purpose, and hit against the ultrasonic sensor when the tribot approach it, but I don't have a right size one at home.

A minor modification to the tribot is to remove the orange color part in the grab area (see picture below the right one), so that it grabs the bottle better.

After having all the fun with the iteration method, I summarize my discoveries here:

1) Ultrasonic sensing does not provide an accurate or sharp data on the border of the object, the accuracy is even worst when detecting distance while the robot is moving.

2) A flat smooth floor is can avoid getting extra noise (although they are always there) and confuse the sensor reading.

3) To run the robot at low speed, you have to use the motor block with motor power control selected. Otherwise, the motor may not move at all. Motor power control is only available on the motor block, but not the move block.

4) The program written detects only one of the edge of the object, and hardcode an angle to further rotate the motors so that it aligns with the center of the object. It is better to memorize the angle of the first edge found, and detect the second edge, then find the mid point of the object. It is difficult to do it now as using the NXT-G a more complicated program runs more slowly on my computer, also, without a compass sensor, just by memorizing the rotations of the motor cannot control it accurately (e.g. you record a +90 deg rotation, then you run a -90 deg rotation it does not go back to the exact orientation).

The NXT-G program is provided here, welcome any comments for improvement.

Alpha Rex

2006-11-11T06:45:00.000-08:00

It took me a whole day to assemble it according to Lego's documentation. As soon as you power up and run Alpha Rex, you might be disappointed by its slow and clumpsy walking movements (we have impressions from Terminator, iRobot...). Anyway, playing the Mindstorm robot set is more than just assembly>download program>press run, there are quite a lot to learn from the design and programming, and below are some of my observations during the assembly/test process:

The assembly instructions could be found inside the NXT software interface, or can be downloaded from Lego Education web site, under the 'NXT Models' tab. The first time when I came across those steps which requires axles, I did not know which one its number refers to, later I found that this was specified at the back cover of the printed manual that came with the Minstorm set.

There are a lot to learn in the walking mechanism of Alpha Rex. Its legs seems to be most complicated in the building process. The purpose of the two motors (namely B and C motors, according to their connection to the output port of NXT) is to control the stride and tilt of both legs.

At first glance, one might misunderstood as if one motor for left leg and another one for the right. Thanks to Jeff Owrey's article "Alpha Rex Avoiding Obstacles" on bNXT which explains the theory on how the toddler-like robot walks: B motor for stride and C motor for tilt. As only one motor is used for both legs, when one leg is moving forward, the other will move backward simultaneously, and vice versa.

As a result, if only the B motor is powered, both feet will touch ground at the same time and it will not walk properly, the C motor then come to rescue, it shifts the weight to one foot and reduce the loading of the other foot which is striding.

In order to walk forward, the robot has to tilt to the left, stride its right foot forward, tilt to the right and then stride its left foot forward, and then repeat the whole cycle. If you look closely on the 'foot' portion, when the leg tilt to the left, the right foot is slightly off the ground. This allows the 'raised' foot to stride forward with less friction, which is the main purpose of the tilt action. Actually, the current design of Alpha Rex does not allow too much raising of the foot from ground, you will still see its outer part a bit touching the ground.

To turn around, say, towards the left, the robot has to tilt the leg to right, stride forward the left foot, tilt back to center and then stride backward the left foot. The actual rotation happens at the last action, where the body is center tilt and both feet strides in opposite direction. This process has to be done several times as only a fraction of turning happens in each cycle.

It is also interesting to see that the current mechanism tilt the body to left when the legs tilt right, as opposed to what we expect (tilting the whole robot to one side). The body is pivoted in the NXT block bottom center and pulled by the linkage on both sides which is connected to the gears of C motor.

In my opinion, this arrangment looks stylish when walking but technically not correct. So, I modified it a bit by connecting the ball joint rods to leg directly. As a result, the Alpha Rex shift its body weight to right when the C motor cause the leg to tilt right. I can see that the striding foot raises more off the ground.

It could be difficult to observe the walking motion when the robot is moving around, the method I used is to switch off the robot and insert an axle part in the motor axis (see picture), then manually rotate it and see how it works. According to Jeff Owrey's article, the motor rotates 270 degrees reverse direction (an arrow pointing downwards means motor reverse rotation) when moving from right to center, another 270 degrees when center to left... etc. You can verify the rotation direction in the help section of NXT software, under the 'motor block' topic.

Another way to observe the legs motion is to reduce the power of the motors, so they rotate slowly and you can study the motions more easily.

When using Jeff Owrey's program to test the walking cycle, it is important to 'center' the walking mechanism before start running. This is because the current robot has no means to reset itself to an initial condition and it does not know where it start from. Without initializing the mechanism, you may end up having the robot walking in a weird way or even in the opposite direction.

To initialize it manually, use the key to rotate the B motor until both feet aligns, which is the center position, and make sure that when you rotate it, hold the NXT body with one hand so that both feet hangs freely from the floor. In the same way, you may have to adjust the C motor manually to center the tilt. For the A motor, you can turn the motor with fingers and center it.

If the robot walks backwards after centering, you will need to turn the B motor with the key one cycle (e.g. turn it until the foot comes back to the same position again).

When assembling the legs, I found a small part which is required in the left leg, it seems that the rubber part is for adding fiction to the leg, but I wonder why the instruction told us to put on left leg only, if you know the reason, please tell me.

Next, we look at the A motor, it powers the rotation of the ultrasonic sensor and both arms. The ultrasonic sensor panned between left and right to check for obstacles. The arms swing by rocking the linkages connected to the gears.

To swing more naturally (one swing forwards while the other backwards), connect the black short linkage in opposite direction. Arms swinging are for visual pleasing only, it is of no functional purpose.

I programmed the Alpha Rex using the subroutines from Jeff Owrey's toddler walking program. The following feature is implemented:

Robot walk and detect obstacle at the same time (not very accurate, but I would prefer it walking continuously instead of periodic start stop).
If obstacle within a certain distance is detected, it walks backwards and turn 90 degs, which direction it turns is randomly generated.
The light sensor is used to detect low ambient light, if light level is low, it will also walk backwards and turn so as to avoid running into dark corners (or under table). I have mounted the light sensor pointing upwards to detect light.
The touch sensor is used to toggle start/stop, it starts the walking process after one press, and when press again (or it bumps into something), it will start the go back and turn process.
The sound sensor feedback 3 levels of output, for ambient (quiet) level, the robot keeps exploring the surrounding, at a noisy level, such as turn up your Hi Fi's speaker volume, it will dance, when the sound level is further increased, it will stop all actions.

The program can be downloaded here, for those blocks that required Jeff Owrey's routine, you can download from bNXT site. Disclaimer: The program may have bugs and Meow the mechanic is not resposible for any damage or injury for using it.

Limitations on the Alpha Rex robot / NXT hardware and software:

Need a rechargable battery pack and transformer. The NXT draw a lot of power and even those high power batteries run out of power quickly, and is very inconvinient to dismount those mechanism on the back of Alpha Rex to replace batteries. If I have the rechargable battery pack and transformer, I can charge it by pluggin a wire instead of disassemble those things.
The mechanism in the feet are so closely packed so that they are rubbing each other while walking, something I do not feel comfortable.
The legs are still touching the floor when strides, so, it is not capable to climb an inclined plane.
As the whole robot is made of plastic Lego technic parts, it is not very rigid, loose parts, backlash and friction between foot and floor introduce errors and thus reduce its accuracies while walking (not walking straight lines).
As we mainly use the ultrasonic sensor for obstacle detection, any obstacle lower than the height of the sensor is not detected, the robot may tumble upon such objects.
While using the NXT visual programming enviornment (NXT-G), I frequently came upon the error "Insane Object at BDHP...." when saving, to avoid corrupting the file, I save each time after I make a change, in case the error comes out again, I revert back to the previous saved file. This is very inconvinient, hope that the bug can be fixed soon. There are some fix you can found here when dealing with NXT-G bugs:
http://forums.nxtasy.org/index.php?s=8a853ebd28ca1a2af5b188b9590c9afe&showtopic=178&st=0
The NXT-G becomes very slow when more blocks are added to the program, this may due to using Flash vector graphics in the interface. Sometimes it just hangs there for a minute (CPU 100% utilization). I hope this can be improved, or I have to look for some other programming environment like Microsoft Robotics Studio. It waste a lot of time waiting, and I think responsiveness is more important then just looking fancy.