Week 9 - September 5-11

Two key components were worked on this week: the IR transceivers and Bluetooth Low Energy (BLE) and RSSI (Received Signal Strength Indicator) troubleshooting and verifying operation.

IR Transceivers

Work began on testing the accuracy of the IR transceivers to see if the direction a signal is coming from can be isolated. The photo shown in Figure 1 is an example of several attempts at distinguishing one side from another that was troublesome for the transceiver.

Figure 1 & Figure 2 - IR transceiver sensor shielding attempt, and IR transceiver when LED is on, meaning the receiver on that side detected the other transceiver

Several reasons were considered why this was occurring, and a few steps were taken to prevent them such as lighting, sensors interfering within each other, and separating them physically by wiring them apart (only one was tested by soldering a wire in between it and the circuit board. The lighting change did not affect anything, nor the isolating by custom wiring on its own. It was noticed, however the transceiver that was connected as intended actually held a better sense of direction to where the other transceiver was, and that was because the receivers each were covered by the board they were soldered on, but direction still was not made as clear as needed. To finally get the right direction on the LED that matched the physical direction, a cross like boundary was put in between the receivers, and that seemed to do the trick, properly cutting off one receiver from the other and allowing the least amount of interference, showing two LED’s blinking only when one was moving from one side to another. The video shown in Figure 3 shows a demonstration of this, while Figure 3 has a close view of how it was set up. The arrow next to the video is where the other transceiver was located when recording the demonstration.

Potentially the final design will have the receivers not only on their own respective side of the cart, but also boxed in such that they only receive a signal when the user is on their side. (WARNING: Please lower or turn off volume for videos!)

Figure 3 - Demonstration of IR transceivers

Figure 4 - Other transceiver with reference to the one rotating

As far as their detection angle goes, the transceivers are pretty sensitive to picking up the IR signal when it is their general direction, so as of now it should not be too much of a concern whether they would be able to see the user if they were to be boxed in for isolation from the other receivers. For the test, one transceiver was connected to a wall plug supplying 8.43V and the transceiver on the cart had a battery pack supplying 7.7V.

With a computer and TV on at the same time, the transceivers also did not show too much distortion seeing the signal when the cross was put in place, as when one transceiver is switched off, the other does not have any LED’s on other than the one indicating it is powered on.

BLE RSSI Troubleshooting

Work on the BLE RSSI tracking continued to help refine and increase the accuracy of the conditions that say whether the cart and the user are getting closer or farther from each other. The code was expanded to specify more clearly what defines getting “Hot” (closer) or “Cold” (farther). It still takes longer than usual for the conditions to update, so the rate at with signal strength may be increased or the conditions themselves may be revised. One problem at the moment is it is hard to tell if none of the conditions are seen by the controller about whether it is where it needs to be, hot or cold. So it is considered at the moment for LED’s to turn OFF when no conditions are met to prevent confusion.

Figure 5 - Mock up cart to test RSSI tracking using Bluetooth

As of right now the program can get the cart where it needs to be, to the target distance. To follow, however the cart takes a few seconds for the signal to gradually drop down to a value that will re-engage the wheels and trigger the following operation again. Either having a different value that will trigger the “following” condition after it has reached the target or increase the RSSI sampling rate when in close proximity needs to be tested. For right now, a short demonstration of how the signal reading shows in real time to the Arduino controller when the user would press "call" on the Red Rover app and how the signal increases or decreases as distance does the same. When the cart is moved toward the phone, signal increases and during the small time that the cart is moved away, the signal gradually decreases.

Figure 6 - Distance between phone and cart being demonstrated

Figure 7 - Plot of signal strength reacting as distance changed in video from Figure 6