Results > Future Work
What would we have done differently?
Given a second chance, our group would have implemented the compass heading feature for the blimp bot. On the blimp itself, the compass module was found already connected to the board near the motors as well as on the blimp schematic diagrams. Also, the blimp board was documented to support I2C modules, allowing for full connectivity. However, time constraints as well as confusing revisions to the specification prevented us from looking into implementing this feature. Another feature our group could have added was an extra video feature given more time and freedom to use non-specified hardware. From the start, our group had looked into the video feature as the UART and wireless receivers supported baud rates much higher that 9600. However, we noticed halfway in the project that implementing video would not be feasible as the bandwidth of the debuggers were limited to a baud rate of 9600, making it impossible to send an image (let alone video) back to the computer within a feasible timeframe. In addition, more consideration would have been given to the overall power consumption of the blimp. For the most part, our group utilized methods and coding practices given to us in the sample code as well as code used in previous labs. However, we did not check to see whether the code used was the most power efficient for running the blimp. Future modifications to the blimp would definitely replace loops with LPM calls and make processes more efficient. A lot more improvements could have also been made to the height system to consider low-power consumption and reading errors. Within our current height control system, the blimp uses the motors constantly to maintain a respective height. As a result, we found that after a certain amount of time, the sensors and motors would not function correctly due to the low amount of power of the battery. Future modifications would include implementing the Kalman Filter and incorporating low power modes to smooth height calculations and utilize power. To further add to this, support for multiple byte packets for activating the sensors and the fans could have been implemented to reduce power consumption (can isolate and control the usage of the fans/sensors) and even speed up communication between the blimp and the user interface. Finally, further improvements could have been made to the steering wheel control schema to account for accurate and realistic control of the blimp.