Donate to Science & Enterprise

S&E on Mastodon

S&E on LinkedIn

S&E on Flipboard

Please share Science & Enterprise

$500K Challenge Seeks A.I. Covid-19 Reopening Models

Synthetic biology

(Gerd Altmann, Pixabay)

17 Nov. 2020. A new challenge from XPrize seeks evidence-based models using artificial intelligence to help governments reopen their economies during the pandemic. The accelerated crowdsourced competition has a total prize purse of $500,000, with the first deadline for team registration set for 8 December 2020.

The Pandemic Response Challenge is a four-month competition to help governments and communities find Covid-19 mitigation strategies that still allow economic activity to safely continue. XPrize, the Los Angeles-based group conducting the challenge, says there’s now a need for more sophisticated solutions with rising expectations for an effective vaccine. Governments, communities, and organizations will be asked to plan for reopening their economies under a range of possible conditions, requiring alternative “what-if” scenarios that optimize public health, while reducing the impact on economic and social life.

The competition asks for teams to develop computer models and systems from a common data source that help regional governments and communities plan for economic reopening. That data source is the Oxford Covid-19 Government Response Tracker, offered by the university’s school of government. Amazon Web Services is providing cloud and computing services to participating teams. The professional services company Cognizant, in Teaneck, New Jersey, is partnering with XPrize on the challenge, offering A.I.-based models from its solutions toolbox called Evolutionary A.I., with an online demo also available.

“Through global collaboration,” says Amir Banifatemi, XPrize’s chief innovation and growth officer in a statement, “the output of this challenge could be a game changer when it comes to using data and A.I. to create a route to recovery via a range of economic scenarios created by the competing teams. With vaccines on the horizon, these predictive and prescriptive models for opening society will continue to be valuable for the foreseeable future.”

The competition’s goal is a prediction model that accurately estimates future daily Covid-19 cases, along with reliable  prediction models for economic activity under various plans or scenarios. Because of the urgent need for solutions, the competition is on an accelerated schedule, with the challenge scheduled to complete and award prizes in February 2021.

The first deadline, for teams to register, is 8 December. Registration for this and other XPrize challenges is done through the group’s Prize Operations Platform, with a $100.00 fee payment required. Competition guidelines can be downloaded from XPrize.

During the challenge’s phase 1, between 8 and 22 December, teams will be asked to write and submit their prediction models. After 22 December, the submitted models will run automatically with their predictions ranked online based on accuracy, until 12 January 2021. From 6 to 12 January, judges will further evaluate the models on quantitative and qualitative criteria.

Based on that initial judging, the top 50 models and phase 2 finalist teams will be announced on 13 January. After 13 January until 3 February, finalist teams will refine their models, for the last round of judging, beginning on 3 February until 25 February. Winners are scheduled to be announced on 26 February.

More from Science & Enterprise:

*     *     *

Comments are closed.