|home > Information for Current Students|
|Mothra Strike Timer|
Athletic performance improvement via coaching and feedback is well established. Performance improves when the athlete is able to measure performance and test strategies to improve it in real time. Karate performance, in this case striking speed, should also be amenable to improvement by feedback.
The Mothra device is a strike timer designed to provide immediate feedback about the speed of a striking technique. Mothra is fundamentally a reflex timer. Users wait for the signal light to illuminate then strike the board, breaking an Infrared (IR) laser beam to signal completion of the strike. Feedback is provided by red and green lights where red indicates the strike was slower than the baseline and green indicates it is faster than the baseline.
Mothra’s striking board – Figure 1 - is a one by five and half foot board with one by three inch side boards glued on the 4 sides of the striking area. The smaller boards provide structural support to the striking board, protect the padding from damage in transport, and for the top and bottom boards, hold the IR laser and detector. The striking surface has a one inch thick, closed cell foam pad and the back has two inches of the same type padding. The back padding protects the wall behind the board and provides spacing for the dojo wall’s baseboard trim. The upper board holds a low power semiconductor IR laser that projects a two inch spot on the bottom board, centered on the detector. The small size of the laser diode made alignment a challenge and epoxy was used to secure the detector and its mounting board in the correct direction and with enough mechanical strength to survive both moving and striking.
Two rows of two-color super-bright Light Emitting Diodes (LED) are attached to the side boards. Lighting red, green, or both for yellow, they provide all signaling to the user.
There are three circuit boards mounted at the base of the board as shown in Figure 2. The first, top most board connects the super bright eight LEDs to the middle “control” board. The control board’s crystal controlled timer supplies a 1.7 KHz clock giving better than one millisecond resolution, its logic routes and controls the feeds to the counter inputs, and its transistor switches control the LEDs. The third – bottom - board is a commercially available hobbyist Universal Serial Bus (USB) interface board that connects to a computer via a 16 foot USB cable. The board has eight digital outputs, five digital inputs, two counter inputs, and two analog inputs which are not used in this design but could be used to monitor the system for laser alignment errors, circuit voltages, etc.
An old Windows (TM) laptop computer to runs the Mothra software. Timing is done on the circuit cards so the speed of the laptop does not affect performance.
The Mothra software is written in Java. When the program starts it checks to insure the correct Windows drivers are installed, verifies that it can talk to the USB board, and tests the Mothra board for proper functioning. The laser is turned on and off and the detector and circuitry are tested to insure they are still in alignment and functioning. The timer and its control circuits are tested by performing sample counts, and these become the baseline for converting the number of clock pulses counted to time in milliseconds.
Each user’s data is stored as a Java object and text file, and when a user logs on to the software their prior data is recalled. If they are a new user they are asked for their age, rank, school, and time in karate. The software also records the user’s correct distance from the device. This distance – along with striking form - must be repeated with each strike for the data to be meaningful. The last five strike times of the prior session are used to calculate a target baseline for the current session. Old baselines are kept and only the smallest baseline is used. New users strike without a baseline and get no feedback, creating an unbiased starting point for measuring improvement.
Five flashes of the LEDs signal the user to prepare and then at a random time between one and four seconds the computer turns on the laser and the red and green LEDs. This also connects the timer to the first counter, feeding it pulses until the laser beam is broken. When the beam is broken the LEDs are turned off and the board’s circuits automatically switch the clock to the second counter until the beam is restored, measuring time on the board, or ‘dwell’ time. When both are complete the software extracts the counts from the USB board and divides by the sample count, converting the number of counts to milliseconds. The LEDs are then lit with green for times shorter than the baseline and red for longer times. The data is recorded and displayed on a graph, along with the baseline. Each user gets 40 strikes per session, broken into sets of ten. Strikes longer than five seconds are not counted and the program can be exited anytime without loss of data.
Repeatability is required for meaningful measurements. Users are positioned far enough away from the device that they can just touch the pad at full extension. A tape measure and small block of wood both measure and keep the distance. The software records this distance but allows it to be changed when appropriate, for example, as children grow the distance needs to change to reflect their larger size. Mothra is not designed for high impact strikes though it can tolerate moderate impact.
All new students are baselined on Mothra after a brief introduction to proper punching form, however,
they will not be allowed to practice with it until their form has matured and become reasonably repeatable. Advanced students train regularly on Mothra.
Improvements can be measured by reduction of variability and striking time. Strike time improvements can be seen in student’s baselines. Mothra always chooses the smaller of either the historical baseline or the baseline from the last session so improving performance will show a decreasing stair step. Figure 3 shows right front punch baselines for three students. Student A improved by 7.2%, Student B improved by 9.2%, and Student C improved by 12.1%.
Figure 3: Right Front Punch Baselines
Repeatability can be measured by the standard deviation where larger standard deviation indicates greater variability – less repeatability – in a student’s response time. Figure 4 shows the standard deviations in right front punch times for the same three students.
Figure 4: Standard Deviations
The standard deviation is calculated in 20-strike windows as is reflected in the width of the square peaks. The square peaks derive from the occasional long punch time students experience due to distractions. For example, in Student C’s performance, at strike 36 the punch time was 590 mS but at strike 37 the punch time was 2,484 mS. In a 20-strike forward window, where to calculate the standard deviation at strike 1 one uses strikes 1 through 20, this anomalous data point will appear at strike 18 and continue through strike 38, as can be seen in Figure 2.
The data indicate an improvement in strike time repeatability. For example, Student A’s standard deviation for the first five strikes is 106.9 but for the last five calculated standard deviations is 41.4, an improvement of 163%.
An unexpected side benefit is the validation of Shorinjiryu form. Advanced users report a tendency to abbreviate form which I believe, having experienced the same thing, is because they are not allowed to strike with full power. The full twisting punch tends to be replaced with an arm-only strike. However, recognizing this and focusing on using twisting form seems to improve strike times. One student’s performance provides clear validation of using proper form.
Figure 5 shows the punching times for Student A’s right and left front punch. Student A is right handed. One can immediately see a difference in the variability between the left and right front punches. The left front punch is significantly less variable than the right. When asked about the difference Student A responded that for the right punch he was trying to just hit fast but for the left he tried to use the best form possible. The average punch times for the first 12 right and left punch strikes are 642.3 mS and 637.8 mS respectively. While this is only a 0.7% difference, the standard deviation drops from 113.7 to 44.3, on par with the improvements noted for the right punch after near to 100 right hand strikes.
Figure 5: Student A Punching Times
Shorininjiryu Karate-do combines the experience of traditional Japanese karate with modern understanding of athletic performance. The Morthra strike timer is just one example of how Morning Star Martial Arts continues to combine these two for the betterment of students.
If you have any questions about Mothra, feel free to email at firstname.lastname@example.org.