Campus RAKSHAK (Protector) A DST (NMiCPS) CDSS for Population Safety
Updated: Mar 20, 2021
Our campuses are reopening. Soon we will have a full-strength of students, faculty, and staff on our campuses. The population in large campuses like the IISc or IIT Bombay could be as high as 11,000-12,000. While the student community may be predominantly residential, we also have staff and vendors move in and out of campuses. The prototype campus in this discussion is a college campus, but in principle, the solutions described below extend to neighbourhoods, communities and technology parks.
Despite the current low incidence rates in many states, some states' surges are of concern. Given the spread due to asymptomatic CoVid-19 carriers and the emergence of new variants, we must be ready with cost-effective population screening and testing strategies, both for CoViD-19 and other future infections. If the virus mutates and negates the gains made with therapeutics and vaccines, screening and testing will become the most important weapons to battle the epidemic.
There is a dire need for a home-grown decision support platform that is customisable for our campuses to serve as an advisory system for our campus administrators. To build such a decision support system, the Coordination Committee of the RAKSHAK project (an initiative under DST- NMICPS} supported the Campus RAKSHAK project in September 2020. Two groups at IISc Bangalore, another group at IIT-Kharagpur, and the group who developed Tapestry Pooled Testing Methodology at IIT-Bombay and GoCoronaGo contact tracing App (both were also supported under RAKSHAK programme) collaborated to build the Campus RAKSHAK platform. The project's goals were:
· to validate a model for keeping campuses safe
· to help campuses feel safe at affordable cost points using indigenous innovations in testing and pooling strategies, and
· to deliver a technology implementable by campuses across the nation and the world, which can help manage future epidemics and evaluate public health interventions.
The Rakshak Coordination Committee is happy to report the creation of a decision platform built on integration of three complementary agent based simulation models (CampuSim (IISc), Campus Rakshak Modeller (IIT-Kgp), EpiSimmer (IISc, RxCoVEA), "Tapestry" pooled testing (IIT-B) algorithms, contact tracing methods using the GoCoronaGo App (IISc) and a badging scheme HealthBadge (HealthBadge.org) driven dashboard for monitoring safety on campuses.
Investigators: C. Bhattacharya (IISc), Rajesh Sundaresan (IISc), Nidhin Koshy IISc), Manoj Gopalkrishnan (IITB), Yogesh Simmhan (IISc), Tarun Rambha (IISc), Adway Mitra (IITKgp), Sudeshna Sarkar (IITKgp), Inavamsi Enaganti (HealthBadge, IISc), Vijay Chandru (IISc, HealthBadge)
The first pilot rollout has been in effect for a fortnight now at the IIIT-H campus in Hyderabad (the director is a computer scientist trained in computer vision at CMU - PJ Narayanan) - here is a blog that describes the Rakshak (translates as protector) pilot https://blogs.iiit.ac.in/monthly_news/when-your-campus-is-a-rakshak/
When your Campus is a RAKSHAK - Action at IIIT-Hyderabad
“The high ability of asymptomatic covid-19 carriers to spread the disease, implies that we would need to be ready with cost-effective population screening for CoViD-19.”
- Dr Charles Cantor, addressing RxCoVEA in May-June 2020
Campus Rakshak - A decision framework for population safety in campuses
Campus Rakshak is a decision framework that integrates various applications and tools into one Rakshak framework. The applications include the Go-Corona-Go app for contact tracing, the Tapestry Pooling tool for campus-wide testing, epidemic simulators, and recommender systems. These form a single platform that monitor, predict, and recommend mitigation strategies for campuses against CoViD-19. We now describe the components of Campus Rakshak.
Go-Corona-Go App: The Go-Corona-Go app uses Bluetooth broadcast and scanning to determine nearby users. It builds an anonymous contact-graph for the institution. If a positive case is detected, and if contact tracing is necessary, the institute authorities have the capability for selective de-anonymisation, which can aid contact tracing. In addition to deployment at IISc, deployment at IIIT-H has started. Ethics and admin approvals, Initial pilot, IT Training, Intranet website, public posters, etc., are completed. BLE Beacons have been purchased and configured. Currently, there are more than 50 BLE beacon installations on the IIITH campus. More deployments are likely as more students arrive.
Tapestry Pooling: Tapestry pooling is a novel smart pooling scheme to test a sample pool using only a few tests quantitatively. The tapestry pooling algorithm assigns individual samples into multiple pools. Based on the results of each pool, the algorithm predicts the CoViD-19 positive samples. Tapestry pooling, combined with swab pooling, can provide up to 100x savings in testing costs. Tapestry pooling is the only Indian entry in the XPRIZE finals for CoViD-19 testing. IIIT-Hyderabad has implemented tapestry pooling on its campus. IIT Bombay will likely implement this soon.
Epidemic Simulators: Three agent-based epidemic simulators, each with complementary capabilities, were developed as part of the Campus Rakshak project.
CampusSim: CampusSim is an agent-based simulator that models the disease's spread via various interaction spaces on the campus. It generates an interaction graph between agents and interaction spaces. Agents are assigned different interactions spaces such as classrooms, hostels, messes, and cafeterias. CampusSim is a parsimonious model that enables a fast and concise simulation of the epidemic spread. The simulator also models complex intervention strategies (case isolation, hostel quarantine, etc.).
Campus Rakshak Modeller: Campus modeller is an agent-based simulator that models the mobility of agents on a campus a little more faithfully than CampusSim. Based on the agents' locations, the interaction graph between agents is updated. As the simulator tracks the spatial movement of agents, it can model spatial interventions, for example, closure of a particular area on campus. The modelling of spatial movements also enables rich visualisation of activity on campuses.
EpiSimmer: Episimmer is a modular and flexible simulator that enables users to specify their custom disease progression models and intervention strategies. Episimmer can model scenarios not supported by the other simulators, such as class scheduling, class staggering, etc. EpiSimmer can be used to develop decision support and recommender systems.
Health Badge is a badging system that enables the real-time implementation of interventions. It uses information from the other components, such as test results, contact graphs, simulation analyses, etc., to prescribe activity levels for everyone on campus. The badging system also suggests the candidates for testing. We have deployed the badging system in the IIIT-Hyderabad campus.