Hack@CEWIT Prize Information: 

Check out Devpost!

Hack@CEWIT Projects 

Ideas For Hackers

Below are projects that our Sponsors and Industry Partners have suggested as potential projects for your Hack.  Some will provide data sets to use, suggestions of hardware, and some will even offer a special prize you can compete for when you submit your projects in Devost Sunday Morning by 8am. Our hope is not to tell you what you should achieve, but instead INSPIRE you - to be creative, take risks, and have the tools to succeed!

Click on the topic, and check out the idea!

• CEWIT Building Energy Management & Sustainability Hack

  

   

  Affiliated Organization: Office of Sustainability at Stony Brook University

  Project Description:

  We are looking for creative, informative, and actionable usages of the Office of Sustainabilities CEWIT building data

  To learn more, Click here!

• Mobileware

  
  

  Workshop Title

  "Driving Towards a Sustainable Future: The Role of Carpools"

  Company: Mobileware

   

  Background:

  onTime Carpool is a first and last-mile on-demand carpool for commuters.  It is supported by NYSERDA

   

  Topics to discuss:
  1. Introduction to Sustainability and the Importance of Carpooling
  

  • Definition of Sustainability and its Importance
  • Impact of Transportation on the Environment
  • Benefits of Carpooling for the Environment, Community, and Individuals

  2. Carpooling and Technology
  

  • Overview of Carpooling Apps and their Features
  • How Technology can Help Boost Carpooling Adoption
  • Examples of Successful Carpooling Programs

  3. Carpool Matching Algorithm
  

  • Overview of Matching Algorithm for Carpools
  • Importance of Gender Matching, Detour Limits and Time Preferences
  • How the Algorithm can be Customized for Different Needs

  4. Building a Sustainable Carpool Culture
  

  • Steps to Encourage Participation in Carpooling Programs
  • Developing Carpooling Policies and Incentives in Organizations
  • Engaging Community Members in Carpooling Programs

  5. Challenges:

  1. Building a social media platform that utilizes the REST API to allow users to share carpooling check-in, experiences and connect with other carpoolers.
  2. Creating a blockchain-based incentive system for carpooling that rewards users with digital tokens or cryptocurrency for participating in carpools, based on factors such as distance traveled, number of rides offered, and overall impact on the environment. The platform could also allow users to donate their rewards to environmental causes or exchange them for discounts at participating businesses.
  3. Developing a web application that uses the REST API to match carpoolers based on their schedules, locations, and preferences.
  4. Designing a game that uses the REST API to simulate carpooling scenarios and encourages players to make sustainable transportation choices.
  5. Creating a tool that leverages the REST API to help universities and students manage carpooling programs and encourage participation.
  6. Developing a matching algorithm for carpools that incorporates real-time traffic data, to optimize route planning and minimize travel time and carbon emissions.
• NASA Challenge

  

  Check out the Data supplied by NASA re: weather - the biggest driver of climate change and enviromental changes.  NASA project scientist Jennifer Wei will be giving a workshop at 8pm Friday evening showing how data can be used to prdict weather patterns.  Following Jennifer will be Thilanka Munasinghe on how to collect and use the data.

   

• NYPA Challenge

  Weather Forecast and Dynamic Line Rating of Overhead Transmission Lines

  brought to you by

  

  Introduction

  In order to achieve New York State's ambitious goal of attaining 70% renewable energy production by 2030 and 100% zero-emission electricity by 2040, we need to plan a significant amount of solar and wind power generation. However, building hundreds of miles of new transmission lines to accommodate the increased power transmission capacity is not feasible due to the difficulty of acquiring large parcels of land. This is where Dynamic Line Rating (DLR) comes into play, as it has the ability to significantly increase the capacity of existing transmission lines.

  The capacity of a transmission line, also known as line ampacity or line rating, is determined by the maximal current that the conductor can carry before its temperature rises to the limit, say, 95 ^oC, due to ohmic heat. Factors such as air temperature, solar irradiance, and wind can affect heat absorption and dissipation, ultimately impacting the line rating.

  Traditionally, conservative static line ratings are used, assuming high temperatures, high solar irradiance, and low wind, to ensure a large safety margin. However, with more reliable real-time field measurements and accurate weather forecasts, we can now push the limits to the extreme using real/forecasted weather data, which is the Dynamic Line Rating (DLR).

  There are two types of DLR: real-time DLR, which uses weather information measured by field sensors to calculate the rating at this moment, and forecasted DLR, which uses weather forecasts to predict the rating in the near future for power market trading.

  However, there are still many challenges be addressed, where we need your help.

  1)       How good is the weather forecast?

  How to measure the accuracy of weather forecasts?  What are the conditions that may affect the accuracy?  How the accuracy changes with the length of the forecast?

  2)       How to convert weather forecasts to local ambient condition of the line?

  Weather forecasts cover large areas.  They normally differ from the local measurements around the line.  The connection between them is nonlinear, multiple-to-multiple, and complicated.  How to convert?

  3)       How to calculate line rating from weather data?

  Yes. IEEE standard 738 specifies the calculation.  But that requires accurate measurement of line properties such as conductor surface conditions, which are normally unavailable.   Can we use regression to find these properties if the weather and rating data are given?  Can we utilize Machine Learning or Deep Learning?

  Hope by diving into these challenges, you will have a taste of a real-world scenario in this hackathon.

  What you can gain from the project

  • Get a feeling of real-world problems. See how they differ from the textbook problem sets.
  • experience data cleaning. You will see how messy the real-world data are.  Sorry we are never in a perfect world.
  • Understand the art of weather forecasting: the potential and limitation, and the performance.
  • Play with Dynamic Line Rating.
  • Enjoy the process and have fun!

  Problem set

  In this project, we focus on analyzing one span of an overhead transmission line, which connects two towers or poles. However, we do not have access to certain information about the line, such as its direction or the size of the conductor. We provide two datasets for analysis: one contains field measurements of the weather from sensors located at the span (Field Measurement.csv), while the other contains weather forecasts for that area (Weather Forecast.csv).  These two data sets are from two independent sources.

  In “Field Measurement.csv”:

  The first column gives the date and time of the measurement.  It covers hourly measurement from “10/1/2021 12:00:00 PM” to “4/30/2022 11:00:00 PM”.  Please notice that many entries are missing due to equipment malfunction or communication problem.  Also the time is in UTC.

  The second column gives the wind speed in m/s.

  The third column gives the wind direction in degree.  The wind from the north is 0^o, and the wind from south is 180^o.

  The fourth column gives the solar irradiance in W/m².

  The fifth column gives the ambient temperature in ^oC.

  The sixth column gives the calculated line rating in Ampere.  All the rating in Apr. 2022 were removed for the challenge.

   

  In “Weather Forecast.csv”, the data covers similar time span and the structure is similar.

  The first column gives the date and time of the weather forecast.  There is only one weather forecast issued per day at 12 noon for the next 84 hours in an hourly base.

  The second column gives the forecast steps in hours from the time in the first column. 

  The third column gives the targeted time slot in the forecast. 

  For example, in row 10, the time to issue the forecast is “10/1/202112:00:00 PM” in the first column.  The forecast step is 8 hours in the second column, and the valid time is 8 hours later at “10/1/2021 8:00:00 PM” in the third column.

  The rest are similar to those in “Field Measurement.csv”.  The rating from the forecast in Apr. 2022 were removed for the challenge.

   

  The requirements are:

  • Clean the data. Evaluate the data quality.  How accountable they are?
  • Evaluate the accuracy of the weather forecast. Conditions and factors may affect forecasts.
  • Convert the forecast to the local measurement so that once we have the global forecast, we know the local forecast.
  • Find the line rating for April 2022 for both local measurement and forecast.

   

  Supporting materials:

  • Field Measurement.csv
  • Weather Forecast.csv
  • IEEE Standard for Calculating the Current-Temperature Relationship of Bare Overhead Conductors.pdf
• Redshred Challenge

  A Metaverse of Documents

  brought to you by:

    

  Who we are and why we are here:

  We are looking for innovative ways to represent and remix information from within documents in AR/VR/MR environments. 

   

  As an API-first, SaaS platform, RedShred empowers developers and data scientists to tailor the way they interact with document-hosted knowledge to build smarter applications. Any and all concepts are welcome as long as they explore unconventional ways to present information from documents in a 3D immersive environment. 

  Project Technologies Used:

  • RedShred APIs (documentation specific to the hackathon available here)
    • API keys will be provided to participants at the event
  • Unity 3D, Open XR, or similar technologies for immersive environments

   

• Vasomedical Challenge

  ECG data signal quality index

  Brought to you by

  
  

  General
  Vasomedical provide ECG and ABP monitoring devices and Cloud-based analysis software
  platform. The projects proposed here will be related to ECG data transferring and analysis.

  Background
  Our ECG analysis software is designed to analyze data with varying lengths ranging from 24
  hours to 30 days. This data is recorded using ECG Holter recorders and may be susceptible to
  motion artifacts and electromagnetic noise from the surrounding environment. As a result, it is
  important for our software to be able to evaluate the overall signal quality and apply
  appropriate pre-processing to the data before the algorithm performs a full analysis to identify
  any arrhythmia (irregular heartbeats) events.

  
  Challenge
  Design an ECG data quality index. The following goals should be considered,
  1. Efficiency: The quality assessment should be completed quickly, within 30 seconds for a
  24-hour long data, or 20 milliseconds for a 15-minute long data. This assumes that the
  data has been loaded into RAM.
  2. Quantifiability: The index should provide a quantifiable measure of the ECG data quality.
  This means that the algorithm should output a numeric value that reflects the quality of
  the data.
  3. Programming language: The algorithm should ideally be written in C++, although other
  programming languages can also be used.

  
  Provided data
  MIT-BIH Noise Stress Test Database, https://physionet.org/content/nstdb/1.0.0/, data format: MIT 16 
  The data format for the rest of three data sets is MIT 212.
  MIT-BIH Normal Sinus Rhythm Database, https://physionet.org/content/nsrdb/1.0.0/ 
  MIT-BIH Arrhythmia Database, https://physionet.org/content/mitdb/1.0.0/ 
  MIT-BIH Long-Term ECG Database, https://physionet.org/content/ltdb/1.0.0/ 
  Note: Not all data need to be used. The way to decipher the data can be found here.

  https://archive.physionet.org/physiotools/wag/signal-5.htm#toc3 
  https://archive.physionet.org/physiotools/wag/header-5.htm 

  
  Tool for viewing ECG waveforms
  Tool Name: WAVE
  Learning resources:
  https://archive.physionet.org/physiotools/wug/wug.htm  and the software download link
  https://archive.physionet.org/physiotools/wug/node85.htm#app:setup 

• Vasomedical Challenge 2

  ECG data compression

  Brought to you by

  

   

  Background

  When ECG data is recorded and stored in an ECG Holter device, it is important to minimize thestorage required for the data. By compressing the data, less storage space is needed, and it takes less time to transfer the data from the device to a computer or to the cloud.
  To achieve efficient data compression, it is important to use compression algorithms that preserve the quality of the ECG signal. Lossless compression algorithms can be used to reduce the amount of data without compromising the signal quality. These algorithms work by identifying and removing redundant information in the ECG signal, resulting in a compressed file that can be stored or transferred more efficiently.
  

  • Challenge
    
    Design an ECG data compression algorithm so it could be either used during ECG data acquisition with a recorder or used on the data stored on PC or Cloud. The following goals should be considered: 
    
    • Efficiency: The quality assessment should be completed quickly, within 30 seconds for a 24-hour long data, or 20 milliseconds for a 15-minute long data. This assumes that the data has been loaded into RAM.
    
    • Effectiveness: Data compression ration should be greater than 1.3, and the higher the better.
    
    • Programming language: The algorithm should ideally be written in C++, although other programming languages can also be used.
    
    

  Provided data
  The data format for the two data sets is MIT 212.
  MIT-BIH ECG Compression Test Database, https://physionet.org/content/cdb/1.0.0/ 
  MIT-BIH Long-Term ECG Database, https://physionet.org/content/ltdb/1.0.0/