No piece of hardware or software is perfect, and the Finch, as a combination of hardware and software, is doubly prone to problems. We have documented all issues we are currently aware of here.
Software Hangs on Connecting to Finch
Description: You run a correctly compiled program and it hangs after "Connecting to Finch". The LED stays in color fading mode.
- Finch is unplugged.
- The USB hub of the computer did not properly register the Finch when it was plugged in. This can be accompanied by a "USB Device not recognized" error.
- The computer went to sleep when the Finch was plugged in and did not properly register Finch upon wakeup, leading to a "USB Device not recognized" error.
Solution: Unplug Finch, count to ten, plug Finch back in.
Program Compiles and Runs, but Nothing Happens
Description: You run a program and the Finch LED turns off, and then turns back on five seconds later.
- A previously run program threw an exception and crashed. This sometimes causes an errant thread that keeps the Finch from properly connecting.
- Running 32-bit Java on a 64-bit platform. If you have a 64-bit OS, the Finch requires that your JDK is also 64-bit.
Solution: If the issue is due to an errant thread, unplugging the Finch and plugging it back in will solve the issue. If it does not, try restarting the IDE. If the problem is the JDK, download a 64-bit JDK.
Description: A program instantiating the video object compiles, but at run-time either java crashes or you get an exception.
- No webcam plugged in or available.
- The third-party library we use for video support (LTI-Civil) does not work in 64-bit environments.
- For a currently unknown reason, LTI-Civil does not work in BlueJ, Greenfoot, Netbeans-Jython, or DrJava. It does work in Eclipse, Netbeans-Java, JCreator, and command line (on 32-bit systems only).
Solution: Use an environment in which video works.
Finch doesn't see an Obstacle
Description: The Finch doesn't seem to detect an obstacle.
Cause: The Finch uses two infrared LEDs and an infrared receiver to detect obstacles. Using infrared is a low-cost and effective way to detect most obstacles, but it is limited:
- Dark and non-reflective objects are harder to detect.
- Narrow vertical objects like chair legs may not be detected.
- Certain black plastics transmit infrared light (they look like glass in the IR spectrum) and so don't reflect sufficiently to be detected.
- On Finches sold before November 15, 2011, interference between the left and right sensors can cause difficulty in sensing objects that are directly in front of the robot. We have instructions for a simple fix to improve the responsiveness of these sensors.
We strongly recommend using lightly colored obstacles like cardboard boxes or white office paper. Covering obstacles in tin foil increases responsiveness significantly.
Finch doesn't move straight/turn accurately
Description: The Finch's movements are not repeatable - a timed turn that moved 90 degrees might move 110 degrees the next time.
Cause: The Finch's tether makes repeatable movements difficult - turns cause the tether to twist, causing a spring force that tends to resist continued turning in the same direction. The tether may also drag on the Finch causing it to turn slightly when it should move straight. You can minimize the effect of the tether by:
- Holding the tether overhead while the Finch is moving.
- Adding weight to the Finch (pennies taped to the underside work well).
- Avoiding activities and assignments that call for accurate movement - the Finch will not do well in maze navigation, but will follow a light source just fine.
Ultimately, inaccurate movement is the main drawback of the tether. The Finch could be wireless, but the cost of a wireless tether + batteries would nearly double the Finch's retail price.
Text to Speech or playing tones doesn't work
Description: Nothing happens when saySomething or playWav is called.
Cause: The Finch class uses the computer speakers to play synthesized speech, wav files, and tones. Make sure your computer speakers are turned on.