Art Krahenbuhl and Dr. Brian Jensen, Mechanical Engineering
The palatometer is a device previously designed by Dr. Sam Fletcher to aid in his work of speech therapy. The device consists of a series of sensors laid out on a grid that is attached to a mold of the user’s palate. As the user talks, the sensors react to the places where the tongue touches and sends a signal through wires to a processing chip, housed in a little black box. This black box is suspended around the user’s neck, and relays the data from the mouth to a computer that creates a digital representation of the sensor’s information on the screen. With this information, speech therapists can analyze the placement of the tongue and help improve pronunciation and speech. The Fletchers had already proven the device effective with test subjects; however, there was a problem holding back large scale use of the palatometer. My research team was put in charge of three main goals in helping to improve the palatometer: make the device wireless, comfortable, and cheaper to produce.
Initially our team began researching various low-power forms of wireless communication. From my experience in information technology, I was aware of Bluetooth’s ability to send information wirelessly with ease. Many hand-held devices such as cell phones and PDA’s currently employ the technology so it seemed like a logical fit. Helping the cause, we found a Bluetooth wireless chip sold by a local business that was small in size and could be interfaced to easily by a serial port. It was a simple choice to make, except there was one large unknown. As the palatometer needed to fit within the mouth entirely, would we still be able to send a reliable signal through the human cheek? Would the human body create too much interference to send a reliable data signal to the computer? To test this unknown, the team bought an inexpensive Bluetooth headset for a cell phone. With the cell phone on, the headset was placed in the mouth and a connection was established. From across a fifteen foot room, we were able to establish a reliable connection and send data from the headset to the phone. Our first obstacle to create the wireless communication was the earliest and easiest hurdle to overcome.
Next, the team began work on creating a more comfortable palatometer. As shown in Figure 1, the device was bulky and required many electrical leads to come out of the mouth. This presented a problem to the users of the device. How was speech to be improved if the device hampered speech itself? Our team decided that the best approach to fixing this issue was to contain all of the sensor and data transmission equipment in the mouth. This shift in design presented the greatest problem to our team. It was now our responsibility to find equipment that was not only small enough to fit in the mouth, but efficient enough to be run off of a three volt battery for at least an hour. As creating a new circuit board layout was outside of our budget, we opted to create a prototype that would act as a proof of concept for our design. After purchasing a microprocessor and a development kit, we faced another set back. There was no one that could program the necessary code to read in the inputs from the sensors, then relay that information to the bluetooth transmitter! Luckily, after a few weeks, we were able to contract a programmer that could provide us with the necessary code. While we didn’t actually create a fully functioning prototype that could fit in the mouth, the components that we used with the development kit could be assembled on a flexible circuit board and placed in the mouth.
Finally, the team needed to make the design affordable so that it could be mass produced. The palatometer is a charitable, non-commercial product. The intended audience could not afford the $1,000 buying price. Unfortunately for the team, electronic equipment that is small and low power, like the equipment we used in our prototype, is expensive and cutting edge. Another problem with creating an electronic device is that prices come down as quantity goes up. However, buying thousands of wireless transmitters is difficult to do, especially when the design might fail. Hence, finding a precise cost for a single palatometer all depended on the quantity ordered.
In the end, creating a palatometer proved more difficult than initially thought. One of the major problems was that our team was composed of mechanical engineers. While engineers, their focus is more on assembly, not on chip design. There were no members of the team that could design and program the chip like we needed.
Another problem with the product was the constant requirement changes that we received throughout the project. The Fletchers came initially with the idea of making the palatometer wireless. However, by the end of the project, it not only needed to be wireless, but it needed to be compact, cheap, and completely redesigned. By the end of the project, we weren’t even sure that we had met all of the requirements as they changed so often.
In the final report given to our sponsor, we recommended that the palatometer be redesigned when the technology is more mature. The costs to create a working prototype with all of the requirements given us would take years of work and thousands of dollars. It was outside of the scope of a single senior project, especially when the seniors had no practical experience in electronic design.
The redesign of the palatometer was neither a success or a failure. We were able to provide our sponsor with valuable research that could further the project. While a finalized product is not available, the research we created gave instructions on paths that could be taken to create a viable product. Much more research will be required before the product is ready to be used by clients, however, our findings show that creating a wireless palatometer is only a matter of funding and time.