Senior Design Senior Design Website

Batteryless sensor nodes are an emerging technology that enables low-cost, low-maintenance, and long-lifetime embedded sensing. Rather than being wired into a continuous power supply or provisioned with a battery, these sensor nodes survive solely on harvesting energy from ambient sources available in their environment, such as RF signals, solar, vibrational, or thermal. As such, they are a challenge to design since these sources are often highly variable and very low power. Our team was tasked with creating a testbed for researchers to use for the battery-less nodes they are developing.

Final stackup of our testbed integrated with the device under test. The 3d printed board is a stand in for a board that the researchers had not been fabricated yet.

Project Details

My team had the task of developing a testbed system for testing multiple batteryless sensor nodes communicating with each other. The goal of our project was to develop both hardware and software to test these nodes. Additionally, we had the task of making a cleaner batterless sensor node hardware or "BOB" node setup of multiple boards that make it a clean stack. Our system was then connected to this BOB node and monitored aspects of it; wired data such as up time due to the Bob nodes not being on all the time. It also monitored the wireless communication between the BOB nodes. All this data was sent from the monitor to a central computer or host to organize the data and made it accessible for researchers. By the end of the year we had a system for ten of these BOB and monitoring board pairs. We sucessfully did a 3 node setup in a network configuration demostrating scalability. The design was scalable beyond this though, with the ability to be applied to a design with hundreds of nodes.

Within this project I had two tasks. I was elected group leader, so I organized the work effort, organized our timeline, managed communication with our client, managed our documentation, and handled submissions for class. Outside of the leadership tasks, I was a member of our software team writing the embedded code that ran on the microcontrollers to monitor the test and communicates.

Throughout the project, I grew in my embedded systems skills as I read through preexisting code, read documentation, design system archtecture, and wrote code. With that work I developed skills with proposing ideas and describing concerns I see. Additionally, I developed skills within leadership specifically leading a small technical group, keeping meetings on topic, and creating agendas. I also gained skills with figuring out a clients requirements and preferences when developing a solution for them.

Definitions

BOB node- sensor node that are batteryless (powered solely by energy harvesting) that wirelessly communicate with other BOB nodes.

Sniffer node- independent node assigned to each BOB node with the purpose of configuring, emulating sensed events, and monitoring the BOB nodes on times and communication. They communicate with each other wireless to aggregate data.

Sink sniffer node- Destination sniffer node that aggregates data from all sniffer nodes and saves it to a host.

Host- Central logging system that gets data from the sink sniffer node and allows researchers to easily view and manipulate data.

Requirements

Have 5 to 10 sniffer, BOB pairs fully constructed by the end of year.

Scalability in design in Sniffer code and communication and hardware for BOB and Sniffer (1 sniffer per BOB).

Mechanical durability of BOB and Sniffer interface and stand-alone unit for testing use.

Sniffers communicate with each other to pass and compare data from BOB.

Central node that cannot lose power and that can be selected from any sniffer recording data taken by sniffers.

Have a Central node system that organizes files from sniffers.

Sniffers only cause trivial effect on BOB's performance. (Sniffers electrically isolated from each other)

Sniffer to Sniffer communication does not interfere with BOB communication.

Sniffers record both wireless data and wired which could be disconnected.

Include spare GPIO ports on the sniffer hardware for future debug.

Configurable to access unused function on BOB and sniffer.

10 Dollars or less per sniffer node.

Week of continuous power for operation of Sniffer.

Implement a system that is easy to move for testing.

Do not require modification of BOB code for Sniffer implementation.

Implement hardware for wireless mass reprogramming of BOB.

Thorough documentation on the project.

Project Details

Our Design is described in our design document below. Our project had complications of multiple PCB designs with RF components and a network for many codes aggregating data on multiple RF bands. We had a microcontroller on each band for each node comunicating over SPI. We ran into many challenges as we developed the project.
Design Document
Final Poster
Final Presentation

Image Gallery

This is a schematic of the test setup prior to our work. Our's expanding on this by having one "Sniffer Node" CC1352 per each BOB node. Additionally, our setup had ten node pairs and could scale further.

The previous test setup was wire heavy, so our setup put all boards in a simple stack. Additionally, our Sniffer nodes were battery powered and communicated test information back to the host wirelessly making the wires unnecessary and each node isolated.