Project Logs

Collapse

• Electronical Controllers 'n' Things

  Statutory Therapy • 08/21/2014 at 01:44 • 0 comments

  Depending on the pressure sensors we decide to go with; a number of different schemes need to be implemented in both software and hardware to read the sensors. For basic resistive sensors this is pretty easy. A simple resistor divider network can read the a voltage change in the sensor. For basic capacitive sensors the scheme is a bit more complicated. My favorite method is an RC oscillator (Usually some kind of simple op amp circuit) and measuring the frequency shift in oscillator. Things get more complicated when the sensor needs to provide a location as well as a force measurement. Making a scheme to read location based force readings is a project in itself and will be gone in greater detail when we go over construction of the force sensors.

  The controller we are using is called the CACU. This stands for "Costumed Actor Combustion Unit" and it was designed to make actors in theater productions glow in various ways (With the (un?)fortunate side effect of setting arts students on fire. Ha. Ha. You probably think I'm kidding). It was made primarily for use by the Baltimore Rock Opera Society (BROS) for use in their productions. It also was junk. The first version had an XBee port, 8x 4A FET drivers (For LEDS) and 8x Triac Drivers (For EL WIre). All of this was powered off an Atmel ATMega644 at 8Mhz. 

  The most recent version of the CACU is actually pictured in this project description. It features the Atmel ATSAMR21E16A  and 8x 4A FET drivers. It also has an RS485 port and wireless (as ATSAMR21E16A has zigbee compatibility). There is also a blue tooth version for street performances and an AVR version for lameness. It is also tiny compared to the previous version at barely 2 inches square.

  Since this is primarily for driving tons of LEDs, it is not ideal for reading sensors. The design can be modified for sensor readings depending on the pressure sensors we decide on.

• What does that flashing skull mean?

  jhr • 08/21/2014 at 01:36 • 0 comments

  Inevitably we want the software to provide real time, actionable data to the user. Instead of simply presenting a wall of information to the user and having them decipher it to make decisions we would like to go a step further and trigger alerts. These alerts will show the area that potentially needs to be inspected for damages.

  An example of this would be having the pressure sensor receive pressure past a certain threshold and duration which would trigger an alert that the area has been hit. Another example would be the accelerometer moving a certain way and then coming to a complete halt which would trigger a collision in a specific area that should be acted upon. Ultimately we would like the software to report that Joe broke his leg roughly the same time that Joe realizes that something is wrong.

  Creating these actionable alerts will take quite a bit of testing and tweaking to determine what is useful and minimizing false positives/negatives. All of this useful data can be stored for later uses and crunched later to provide other performance reporting.

• Connection thoughts (Wireless woes?)

  jhr • 08/21/2014 at 01:24 • 0 comments

  We are looking at different types of technologies to connect the suits to the client devices. The wireless technology we use will dictated by the environment we are expecting. The number of users, number of devices, the amount of data we are pushing over the connection, etc.. are all variables that we need to consider to select the optimum technology that we will utilize.

  Bluetooth to mobile - Tons of different mobile devices are available and it is easy to connect to these devices. Bluetooth is unfortunately limited in the number of simultaneous devices and may make scalability past a certain point difficult.

  Wifi to computer - 802.11 connection to software running on a PC or mobile device. This wireless technology is quite widely available which makes deployment possible to a wider audience. We currently have nothing developed for this yet but we want to seriously consider this.

  Xbee - is another common wireless technology we are be looking at using. We already have xbee working on the controller which has been used successfully in previous projects. Unfortunately this wireless will require more hardware as the technology isn't widely used in most consumer equipment.

• Pressure sensor selection

  Statutory Therapy • 08/21/2014 at 01:10 • 0 comments

  While there are a crap ton of different pressure sensors out there, which ones are appropriate for integration into the slip suit? There are a number of requirements that have to be taken into account:

  • The sensors have to cover a large surface area.
  • The sensors have to be able to locate on the sensor where a force occurred.
  • The sensor has to respond fast enough to register an injurious impulse.

  While these are pretty vague; they do not take into account sensor processing or possible power requirements or other details necessary for implementation, it's a good basis to begin selection (or at least rejection) of sensors. 

  A set of sensors that we cannot use are any that rigid surfaces for pressures to be 'sensed'  (i.e. rely on the response of the material to hard backing). This is would be a problem with the sensor suit as humans are notoriously squishy and anything that requires rigidity would not work. Since this suit is for sports we cannot impose some sort of rigid backing as that may decrease the mobility of the user. 

  Another issue is that some of the sensors are may not be durable enough or fast enough.  Felt sensors strike me as particularly delicate and probably are not the best idea for sensing hard blows. I am also unsure of their response time since quick impulses that cause injuries may be too quick for the sensors to pick up. 

  Other sensors such as PZTs are extraordinarily rigid and pretty tiny. Since our idea requires sensors that cover a large area, these probably arn't the best solution.

  Two sensor types I'm leaning towards are resistive and capacitive sensing though use of polymer foams and conductive polymer sheets.  For capacitive sensors we are going to take a dielectric foam and attach two conductive electrodes to the ends. The squish of the capacitor will provide us with a force measurement. Additionally Zoflex conductive sheets will provide some resistive pressure sensing. Localizing these sensors can be done with some intelligent design of the electrodes and  making multiple arrays.

  Here is a video I made of one of these capacitive sensors working with a simple ATMega circuit a few years back:

  https://www.youtube.com/watch?v=qJrnLTnaMTs

  Additionally we'll probably implement some techniques discribed in  "Soft Wearable Motion Sensing Suit for Lower Limb Biomechanics Measurements" [ICRA 2013] 

View all 4 project logs