With two others, I researched, advised and piloted a platform that improved Bicycles Against Poverty (BAP) data collection efforts from handwritten to digital input.
• Literature Review
• Evaluation of Solutions
• Report writing
• Form creation (using xml)
Project Impact: Improved Operational Efficiencies & Increased Data Analysis
1. Data collection is more efficient. BAP is able to digitize their data collection efforts at-point-of-collection. This saves them time from having to transcribe it to a digital spreasheet later.
2. Enables deeper understanding of their constituents. They now collect other data beyond repayment transactions.
3. Enables more complex data analysis. For example, BAP was able to see that one village was delayed for a month in repayments and honed on this problem. They later discovered that the village had been devastated by a fire. This pattern would not have been so quickly identified with their old hand written system.
1. LITERATURE REVIEW Examination of what has been done before to digitize data. Two approaches to digitizing data were examined: after it has been collected manually and at point-of-collection.
2. EVALUATION OF MOST PROMISING SOLUTIONS
We examined Magpi, Cybertracker, and various commerical OCR solutions like ABBYY FineReader Express.
Key Findings & Recommendations
OCR Not Accurate
As the above shows,Optical Character Recognition lacks accuracy in converting written text to digital. Existing OCR software used by banks are too expensive for BAP.
Paper is fail safe
Handwritten documentation is still the dominant way to record information in developing regions. It works through electrical failures and when there is limited or no internet.
Use Open Data Kit
ODK is open-source and scalable. It allows for the creation and deployment of customized forms that can be used without the need for internet connection.