Author: ntarakad

This week we spent a lot of time preparing for our design review presentation. We finalized a lot of details that we had spent the majority of last week researching. Now that the designs have been primarily finalized, we have placed orders for the Zigbee shields, ESP32-C3 development boards, displays, and batteries. We decided on the approach to just purchase 2 Xbee hats in order to show that in the event of a wifi outage, we have a plan for how we can rely on a source other than the local network which we can explain how to scale up. While presentations lasted two days and we didn’t have that class time, we all met outside of class to work on the hardware we already had, research new parts that had gone out of stock since we requested them, and send out new orders. 

This past week I started to plan out what should be included in the design report based on the provided guidelines. I went through a lot of different examples of reports from previous years and figured out what is most relevant and necessary to include regarding our specific project. For example, from the design presentation we learned that because we essentially have two MVPs: one for the node with the LEDs and another for the node with the display. With this information, we need to ensure that our design report goes into depth for the components that make up these different kinds of nodes. Furthermore, I think a good portion of our report is going to be dedicated to our design choices and how we got to our final design. This is because of all of the research and considerations last week that heavily influenced the design decisions we made for our presentation. With these in mind, I have a layout that I think makes the most sense to follow for our report. I believe I am on track with regards to our schedule and the next pressing item would be figuring out how to output to the display. While the displays have not come in yet that would be contingent on their arrival which I will touch on for my deliverables.

Regarding my deliverables for next week, on Friday, the ESP32s came in as well as the batteries and early next week we plan to see if we can interface with them. From the previous week we were able to test our temperature sensors with an Arduino, but now that we have our ESP32s it will be important to be able to interface with them and ensure that we know how to read and write data from our sensor readings. I anticipate there to be a learning curve but I am hoping that because of our extensive research into picking these boards we were able to prioritize the interfacing capability so that there would be documentation that we could follow. Ideally, our displays would arrive sometime before spring break and we can work on being able to display information that we program on the board. I anticipate this learning curve to be big as well despite choosing displays that are compatible with ESP32 so we would want to ensure that we plan accordingly while still being on track to complete our project. Regarding the nodes with the LEDs, that would depend on the output from our pathfinding algorithm as well but now that we have our ESP32s, we should be able to test that we are able to drive certain LEDs and not others instead of just testing with the Arduino like we had done last week. With these action items in mind, we will also be working on completing our design review report as the deadline is by the end of the week.

This week I spent a lot of time meeting with Kaashvi and the rest of our team to try and finalize the hardware that we want to order. This is because originally we were all under the assumption that the nodes would be made up of arduinos. After looking at the inventory, we realized that if we wanted to use arduinos, it would get really expensive really quickly. So, we decided to branch out to using other microcontrollers and we decided that NodeMCU ESP32 was our best option. Once we figured out this part, we started to move on towards finding the exact Zigbee shields that we needed after learning that these are actually quite expensive as well. Then we started to look for development boards that had Zigbee capabilities included and realized that either they were not released to the public just yet or they came from unreliable sources. As of now, we think our best option would be to use regular NodeMCU ESP32 development boards and buy a couple Zigbee shields just to prove that in the event that the wifi connection went out, we have a working solution for how to get around it when it comes to scaling up for our demo. We put in an order and hope to get them reviewed and sent out as soon as possible. Once we have the microcontrollers we also finalized decisions on the displays which we sent out as well. We wanted to ensure the displays we picked were big enough for users to read while being in a hurry to get out of the building but also compatible with the ESP32s and shipping from a reliable source. Trying to get a balance of these three components was a bit tricky but we think we found a good selection of displays and sent those out to be ordered as well. Further, we are still waiting on our smoke sensors as they were shipped from ebay which will take until March 1, but we were thinking of finding smoke sensors from a more reliable distributor so we can get and test them as soon as possible. Finally, our temperature sensors came in this week and on Friday we were able to use an arduino just to test its capabilities and thresholds using a lighter. We kept note of the delay of recognizing the temperature and how long it stays at a certain threshold before cooling down. We also added an LED to detect certain ranges and turn on/off at different cutoffs. All of these considerations will be present in our design review presentation.

According to the Gantt chart, my progress seems to be on track. It has been a little frustrating not getting the parts completely ordered but I’m glad we were at least able to test the temperature sensor as it aligns with our schedule. Ideally we could’ve tested smoke sensors as well and instead of using an arduino we could use the ESP32s but we only made that decision later in the week so we tried to keep up with our schedule in this way. We were also able to scope the battery requirements and send out a request for the batteries we want to use for each node which is on track with what we had planned. But those orders we placed will likely only go out next week. All in all, I am pretty much on track with what we had proposed but we anticipated a quicker turnaround time for all of the parts that make up the structure of our nodes.

In the coming week, I want to be able to find a good smoke sensor from a reliable source and start testing that immediately. I also want to figure out how we need to wire the battery to the rest of the node so we can be able to recharge it in the event of a power outage and allow this to be the source of power when necessary. Once we figure out how these circuits are configured and the layout of the ESP32s, we can utilize our Eagle research and plan how we want to design and fabricate our PCBs. We also know that our design report is coming up so we want to make sure that we have a finalized plan and can explain any design choices that we make and show the progress we have made towards those plans.

This week, the ECE courses I have been working on the most have been circuits as well as some software systems with the arduino/microcontroller programming (specific courses would be 18-220, 18-349). These courses aided in support for diodes to recharge batteries, arduinos to read data from sensors and power LED, and general decision making and design planning considerations regarding the research we have been involved with this week. Outside of these courses, we spent a lot of time researching online to learn about what is going to work best with our design (ie what sensors/displays are compatible with certain microcontroller, what batteries fit the design requirements, etc.).

This week, I gave our proposal presentation after all of us put together the slides where we made it a goal to focus on exactly where our use case requirements were coming from and how to best address them. While presentations lasted two days and we didn’t have that class time, Jason and I met after class to plan out how we should progress based on our schedule. We quickly met with Kaashvi to determine when we can order our parts as we planned to do so sometime this weekend or Monday. 

In regards to my own research, I started with the sensor research as they are hardware components that make up our node structure and we need to order these parts as soon as possible so we can start reading from them and testing the nodes individually. As for smoke sensors, I was doing research based on Arduino fire detection methods and came to the conclusion that the MQ-02 smoke sensor might be a good bet as it is low cost, sensitive to flammable gas and smoke, and available for different applications. Furthermore, I was debating between DHT-22 temperature sensor (low cost digital temperature and humidity sensor that measures surrounding air but it requires careful timing to grab data) and the DS18B20 waterproof temperature sensor which have good accuracy and works within the temperature range we want while being easily available to connect to an Arduino digital input. I definitely want to get some more feedback before placing the orders and finalizing the sensors. I was also looking into flame detection sensors which could be useful too. Ideally we can go over these choices with Kaashvi and Tamal early next week before putting in the order. 

In the coming week, we would want to ensure that we order the necessary sensors and hardware that make up our nodes. Furthermore, while we wait for them to come in we can do more research on Eagle so that once we start initially breadboarding our nodes, we can seamlessly use Eagle to prep for our PCB fabrication. I also want to do some research on the battery requirements so we have a more scoped out idea of how we will be addressing them. We also know that the design presentation is coming up and we can start planning out the slides and fixing any details that need to be considered from questions that were brought up in the proposal presentation. In regards to our schedule, I am on track with respect to the research but a bit behind with respect to ordering the sensors. We plan to get out on order on Monday so we can get back on track.