Call for Design Contest
Calling all designers, innovators, and problem solvers! We're thrilled to announce a design contest held by Microchip during VDAT-2024 hosted by Vellore Institute of Technology, Vellore to provide a platform to showcase innovative hardware and software solutions to real-world problems. Whether it's enhancing efficiency, promoting sustainability, improving accessibility, or addressing societal challenges, we want to see your creative ideas come to life.
This contest is open to individuals and teams from all backgrounds and expertise levels. Whether you're a seasoned professional or a passionate newcomer, your unique perspective and innovative thinking are valued.
Track 1: Hardware Platform: Smart, Connected, and Secure Systems for a Sustainable Society
It is expected to involve a hardware/software co-designed solution using either or both of the below family of Micorchip's 32-bit microcontroller units (MCUs).
- PIC32CM Lx Family of MCUs - Robust Security, Ultra-Low Power and Enhanced Touch
- PIC32CZ CA Family of MCUs - High-Performance MCUs With Connectivity Peripherals and an Embedded Hardware Security Module (HSM)
You will be showcasing design approaches and solutions in the areas/domains of AI/ML, IoT, Industrial Automation, Automotive, Medical, Environment & Clean Energy, etc.
Hardware Platform
The hardware platform will be based on Microchip Technology one of the below
Idea Implementation
Implementation should be using kits provided by Microchip Technology. Experts from the Microchip team will guide and mentor during the time of implementation.
Track 2: Optimized AI Face Detection on SAMA5D27 WLSOM1-EK
Design an AI-based face detection system using the SAMA5D27 WLSOM1-EK board and a USB camera. The system should be capable of detecting faces in real-time with high accuracy and low latency(less than a second). Participants can choose to use existing open-source face detection models optimized for TensorFlow Lite (TFLite) or create their own lightweight TFLite models specifically tailored for edge devices.
Key Requirements:
The hardware platform will be based on Microchip Technology one of the below
- i. Hardware Setup: Utilize the SAMA5D27 WLSOM1-EK board and a USB camera as the input source for face detection.
- ii. Software Development: Develop or integrate a lightweight TensorFlow Lite (TFLite) model for face detection. The model should be optimized for edge devices to ensure fast inference times and minimal resource consumption.
- iii. Real-time Detection: Implement real-time face detection capabilities, where the system can detect and locate faces accurately and quickly from the camera feed.
- iv. Accuracy and Robustness: Ensure that the face detection system achieves high accuracy in detecting faces under various lighting conditions, angles, and facial expressions. The system should be robust enough to handle different scenarios effectively.
- v. Performance Metrics: Evaluate the performance of the face detection system based on metrics such as detection accuracy, inference speed, and resource utilization (e.g., CPU usage, memory footprint).
Additional Challenge:
Added value for implementations that meet the following timing criteria:
- i. Inference time < 1.3 seconds
- ii. Capture time < 2 seconds
- iii. Processing time < 0.2 seconds
Hardware Required:
Added value for implementations that meet the following timing criteria:
- a. SAMA5D27 WLSOM1-EK
- b. USB Camera (any vendor)
Track 3: FocusFlow: A Wearable to Manage Student Stress and Attention
Excessive screen time is a growing concern linked to increased stress and student attention difficulties. This problem may be particularly relevant in India, where mental health resources are scarce. We propose FocusFlow, a novel wearable device designed to help students (ages 8-19) manage stress and improve focus. FocusFlow uses sensors to monitor physiological data (heart rate, respiration, etc.) and eye movement. Using AI and machine learning, FocusFlow identifies patterns associated with stress and attention lapses.
For Students: Gentle alerts and real-time biofeedback help them become aware of their state and take corrective actions. It will also provide focus prompts during study time and reminders for meditation and breathing exercises for dealing with stress.
For Caregivers: Discreet alerts with relevant data empower parents and educators to provide targeted support. FocusFlow can potentially improve student well-being, learning, and classroom engagement.
The FocusFlow will implement security features to protect the data from tampering and misuse. The device sends the data over a secure wireless medium to a cloud platform for analysis and provides targeted support and therapies.
Hardware Platform
Track -4: 'Smart, Connected and Secure Systems for a Sustainable Society'
It is expected to involve a hardware / software co-designed solution using Microchip PolarFire® SoC FPGA, making use of hardware acceleration. You will be showcasing design approaches and solutions in the areas / domains of AI/ML, IoT, Factory & Industrial Automation, Automotive, Medical, Environment & Clean Energy, etc.
Hardware Platform
The hardware platform will be based on Microchip Technology PolarFire® SoC Discovery Kit that hosts a RISC-V based PolarFire® SoC FPGA.
Following the first round of screening, shortlisted teams will convert their ideas into applications / solutions using the PolarFire® SoC Discovery Kit. Hardware kits will be shipped to the selected teams.
Idea Implementation
Implementation should be on the Microchip Technology PolarFire® SoC Discovery Kit provided. Experts from Microchip team will guide and mentor during the time of implementation.
Track -5: Embedded System Design 'LoRaWAN IoT Cloud system for moisture management and reporting'
Problem Statement/Design Challenge
LoRaWAN IoT Cloud system for moisture management and reporting
Brief about the system
The system in design is a soil moisture sensor and irrigation system. Since majority of the farmers are dependent on the rain fed rivers or on the ground water, it is more that important to make efforts to save water. Moisture detectors help farmers monitor soil moisture levels, allowing them to optimize irrigation practices. By providing real-time data on soil moisture, farmers can ensure that crops receive the right amount of water at the right time, minimizing water waste and reducing the risk of overwatering or underwatering. This is also a big step towards environmental sustenance.
The architecture of this system is as follows:
Multiple sensors must be placed in the field, to detect the moisture level of each patch of land. However, as a proof of concept, one sensor can be used to demonstrate the capability of the system. The Moisture Sensor sends the data to the RN2483 Board, sends the data over LoRa radio to the SAMR34 board with Ethernet connectivity. This data is sent over the ethernet to the cloud database with proper timestamp. The application running on the cloud uses this information to control the pump, connected to another RN2843 LoRa node. The application on the cloud also presents data to the user through a web page or mobile phone application. This application facilitates users to configure the preset moisture level to control the pump through the web or mobile phone application.
System Modules
- 1. SAMR34 XPLAINED PRO EVALUATION KIT (Part Number: DM320111)
https://www.microchip.com/en-us/development-tool/dm320111
The SAM R34 Xplained Pro is a hardware platform designed to evaluate the SAM R34 family of LoRa® devices.
This FCC, ISED and RED certified board is not only an evaluation platform but also an excellent reference design for developing SAMR34 based LoRa end-node applications.
This kit is supported by the Atmel Studio, an integrated development platform, which provides predefined application examples. The kit also provides easy access to various features of the ATSAMR34J18B device and offers additional peripherals to extend the features of the board and ease the development of custom designs. - 2. RN2483 LORA(TM) TECHNOLOGY MOTE (Part Number: DM164138)
https://www.microchip.com/en-us/development-tool/dm164138
The RN2483 LoRa® Mote is a LoRaWAN™ Class A end-device based on the RN2483 LoRa modem. As a standalone battery-powered node, the Mote provides a convenient platform to quickly demonstrate the long-range capabilities of the modem, as well as to verify inter-operability when connecting to LoRaWAN v1.0 compliant gateways and infrastructure.
The Mote includes light and temperature sensors to generate data, which are transmitted either on a fixed schedule or initiated by a button-press. An LCD display provides feedback on connection status, sensor values and downlink data or acknowledgements. A standard USB interface is provided for connection to a host computer, providing a bridge to the UART interface of the RN2483 modem. As with all Microchip RN family of products, this enables rapid setup and control of the on-board LoRaWAN protocol stack using the high level ASCII command set. - 3. ENC28J60 10Base-T Ethernet Controller with SPI Interface
https://robokits.co.in/development-board/motor-control-boards/interface-boards/ethernet-to-spi-interface-board-enc28j60
The ENC28J60 Ethernet Module utilizes the new Microchip ENC28J60 Stand-Alone Ethernet Controller IC featuring a host of features to handle most of the network protocol requirements. The board connects directly to most microcontrollers with a standard SPI interface with a transfer speed of up to 20MHz.
Target applications include VoIP, Industrial Automation, Building Automation, Home Control, Security and Instrumentation. - 4. Humidity Sensor
Preferably this one:
https://www.industrybuying.com/temperature-and-humidity-sensors-techdelivers-IND.TEM.120691245/?q=humidity+sensors+&cat=
or
Pick one from the link below:
https://www.industrybuying.com/search/?q=humidity+sensors+&cat=
The TECHDELIVERS Soil Hygrometer Humidity Detection Moisture Sensor, a dynamic duo of sensors that provide crucial environmental data for a wide range of applications. They measure the ambient temperature of the surrounding environment. They use various technologies like thermocouples, thermistors, and integrated circuit temperature sensors to accurately detect temperature changes. This data is vital in industries like HVAC (heating, ventilation, and air conditioning) for regulating indoor climate, in manufacturing processes to ensure specific temperature requirements, and in weather forecasting to monitor atmospheric changes. - 5. Google Cloud -
Click Here
With Google Could deployment the data can be accessed from anywhere and the data can be logged on the Server too to further enhance the watering pattern. This data can also be fed into an AI Model (using TensorFlow), to make informed decisions about keeping the farms adequately moist
Getting started aids
- i. SAMR34 + ECC608a Secure LoRa End Node Software Code Example - CLICK HERE
- ii. Advanced Software Framework v3.52.0 - CLICK HERE
- iii. LoRa Technology Mote (LCD version) Source Code - CLICK HERE
Idea Submission
- Your idea submission should cover a description of the idea along with a block diagram of your idea
- Benefits and value addition of your solution, example application scenarios.
- Your team description along with your contact information with mailing address.
- Faculty can be part, but lead author should be student. Presenter in the conference has to be student.
Steps for submission
- 1. Use the design contest submission template provided.
- 2. Fill in all required details with a minimum font size: 10 pt, Times New Roman font. Max two pages including tables, diagrams, and references (if any).
- 3. Submit a PDF version using the CMT portal.
Eligibility
- Participants should be full-time undergraduate/postgraduate/Phd. students from a reputed institute.
