Improving pandemic response, citizen services, and assessing beehive health: The latest from AWS Cloud Innovation Centers

Cloud Innovation Centers (CICs) powered by Amazon Web Services (AWS) aim to empower public sector organizations to quickly create and test new ideas using Amazon’s innovation methodology. With the CIC program, students and researchers, along with AWS teams, focus on solving real-life societal challenges facing the public sector. Learn more about some of the digital solutions on challenges the CIC team published over the last quarter such as working to prevent opioid overdose, discovering new coronaviruses, and using machine learning to monitor beehive health.

Improving citizen service delivery, opioid crisis response, and transportation policy

In Arizona, the impact of the pandemic increased the influx of calls to service providers for information on crisis, referral, and data services. Arizona 211, a non-emergency citizen services hotline, had been fielding approximately 30,000 calls a day, 50% of them related to housing shelter and utility assistance. The Arizona State University (ASU) CIC built a chatbot for Arizona 211 that would guide people through a series of questions like zip code and residential status to provide customized eviction prevention guidance and information. The prototype has since been converted to a production solution and is featured on the Arizona 211 website.

During the pandemic, the opioid epidemic worsened considerably in San Luis Obispo County with opioid deaths increasing more than 31% from the last year. The California Opioid Coalition has committed to the goal of making Naloxone, a fast acting and life-saving opioid reversal medicine, understood and accessible. The Cal Poly DxHub collaborated with the Opioid Safety Coalition in developing a website that provides one-click mail delivery of a Naloxone kit directly to any citizen as well as training on how to recognize and treat an overdose.

The DxHub also collaborated with the World Bank’s Transport Group to examine variations in mobility patterns and public transportation use by gender in Nairobi, Kenya. A student team analyzed transportation and economic data provided by the World Bank to assess potential for data informed policy opportunities for the Kenyan government. The analysis included data about how increasing access to transportation for women could increase gross domestic product (GDP).

In Germany, the Munich University of Applied Sciences manages project KonTEXT, a program for youth involved in the juvenile justice system to encourage reading books of their choice. Studies have shown that educational support can encourage youth to turn away from crime. However, being ordered to read books was often perceived as punishment. The Munich University of Applied Sciences (MUAS) Digital Transformation Lab (DT Lab) developed ReadUp, a chatbot with podcasts and videos designed to inspire and reward youth for reading. The app was developed using various AWS services including Amazon Relational Database Service (Amazon RDS), AWS Elastic Beanstalk, and AWS Amplify, and is accessible via any web browser.

Improving healthcare technology and preparing a pandemic response

The University of British Columbia Community (UBC) Health and Wellbeing CIC published Phase 3 of AI Model for COVID-19 CT Diagnostics and Prognosis. The beta model has been trained with more data using additional confirmed COVID cases. This resulted in a significant change in how the predictions are shown. It’s no longer just a binary prediction of “Yes this is COVID” and “No this is normal lung”; the Beta model also makes predictions on the likelihood of infection.

UBC CIC also teamed up with the UBC Department of Medicine to create a prototype enabling clinicians to efficiently prescribe antimicrobials. This prototype enables clinicians to efficiently view multiple data points for an integrated assessment on what antimicrobial therapy to prescribe for a given infection. This data is visualized in a dashboard that allows the clinician to make efficient and accurate clinical decisions with an improved user experience.

To specifically address the ongoing COVID-19 pandemic, UBC CIC started an international collaboration between scientists, led by UBC researchers, to develop Serratus. Serratus is an AWS-backed cloud computing architecture, capable of analyzing sequencing data at the planetary-scale. With Serratus, the team re-analyzed 5.7 million sequencing samples to identify 130,000+ novel species of RNA viruses. This increases the total number of known RNA viruses by over an order of magnitude, and it was completed in 11 days. This dataset was named “The Open Virome” to emphasize that it is freely available to all researchers and meant to be used to transform the field of computational virology.

Understanding beehive health with machine learning

Bees are essential to agriculture and are the world’s most important pollinator of food crops. To understand the types of environmental conditions that negatively impact beehive health, the Munich University of Applied Sciences (MUAS) Digital Transformation Lab (DT Lab) published a machine learning model to study the movement of bees in and out of beehives. Using infrared images collected from beehives, students used AWS SageMaker Ground Truth to annotate the data to train and improve the accuracy of the machine learning model.

Any public sector organization, including nonprofits, government agencies, healthcare organizations, and education institutions that are facing a challenge can apply to work with a Cloud Innovation Center to help identify new approaches to problems, leverage leading-edge technology, and explore opportunities to better deliver on their mission. Send the CIC team a message if you are interested in working on a challenge to support your digital innovation initiatives and advance your organization’s mission.