Any Questions? /+91 9958826967, 9958294810, +91 11-41644377

For registration call @ 9958294810 or mail at | LAW OPTIONAL COURSE for CIVIL SERVICES MAINS 2021 with "Indian Polity of G.S. Prelims & Mains" Live classes Starting from 1st December 2020. |

Proportionality of Aarogya Setu

Proportionality of Aarogya Setu


The AarogyaSetu app developed by the government appears to be commendable use of technology to provide efficient solutions.

This editorial discusses the proportionality of technological-solutionism vis-à-vis inadequate privacy protection.


2.1 Transmission of COVID-19

  • There are two ways of transmission of COVID-19 according to the virologists viz. :
    • direct person-to-person transmission by inhalation of droplets or aerosols carrying the virus
    • inadvertently picking up droplets from contaminated surfaces
  • For direct person-to-person transmission of COVID-19, the precise relation between risk of infection and proximity is not clearly modelled but there is a consensus that the person-to-person distance should approximately be within 2m for sufficient virus load inhalation.
  • For picking up droplets from contaminated surface, it is known that the viruses can survive for different time-periods on different surfaces, particularly on hard metallic surfaces, ranging from several hours to even days.

As Centre 'mandates' download of Aarogya Setu app, a debate ensues ...2.2 Need of Analysis

  • The government has excessively pushed AarogyaSetu and hailed it as the major instrument in India's fight against COVID-19.
  • Several experts and technocrats have expressed serious concerns about privacy and trust issues in an app based approach (AarogyaSetu).
  • There needs to be a careful analysis of the effectiveness and utility of the app and the balance of the two and it becomes all the more important because there are no publically available detailed and credible evaluation of the efficacy of AarogyaSetu. 


3.1 Working of AarogyaSetu

  • Electronic risk assessment of COVID-19 like the AarogyaSetu uses two main principles
    1. GPS based geolocation
    2. Bluetooth based proximity sensing

3.2 Assessment using Global Positioning System (GPS)

  • GPS is often unavailable indoors.
  • Even outdoors in dense metropolitan areas average unavailability of GPS ranges in 30-40%.
  • Even during its availability, the GPS can have errors to the tune of several tens of metres on a consistent basis.

3.3 Reliability of Global Positioning System (GPS)

  • Hence, for assessment within 2m person-to-person direct transmissions, especially in dense gatherings, the GPS is clearly not a reliable instrument, especially in dense gatherings.
  • Using GPS if everybody within a few meters of an infected individual will be declared infected, it will generate too many false positives.
  • In addition, for a cautious and COVID-19 aware person taking all safety precautions, mere colocation does not necessarily imply high risk of contracting the infection.
  • Hence, GPS may overestimate risks for direct transmissions.
  • Similarly, GPS is also unreliable for indirect transmissions as the proximity with a potential infected indoor surface is most likely to be missed entirely, leading to false negatives.

3.4 Assessment using Bluetooth based proximity sensing

  • For this method of assessment, each device transmits low energy radio beacons isotopically in all directions at periodic intervals.
  • The listening device picks the signal establishing a communication channel between the two devices.
  • The distance between the two devices is estimated on the basis of the strength of the received signal.

3.5 Reliability of Bluetooth based proximity sensing

  • An optimum effective interval rate of radio transmission for effective risk assessment of direct person-to-person infections is not clear.
  • While excessive frequent transmissions will drain out batteries, too wide time gaps in radio transmission on the other hand will lead to false negatives.
  • Another concern is generation of too many false positives.
  • Bluetooth based proximity sensing can overestimate the risk because radio transmissions can establish connections even across large distances in open spaces and across walls, which the radio transmission can penetrate but the virus cannot. This is a major drawback.
  • False negatives are also possible while assessment through Bluetooth due to weakening of radio signals through human bodies like in case when victim carries the phone in the front pocket while the infected person is present in close proximity behind the victim.
  • The Bluetooth based proximity sensing is also ineffective for assessing indirect transmission of infection. The corona virus can survive on contaminated surfaces for hours or even days hence for effective assessment the intersection of smartphone trajectories will need to be computed not only in space but also over large temporal windows.
  • For this assessment, the Bluetooth based proximity sensing which are isolated communication events over narrow temporal windows between two smartphones will be rendered ineffective.


  • The privacy aspects in AarogyaSetu app have also not been effectively implemented.
  • AarogyaSetu uses a static transmission id for every smartphone which is fixed at the time of registration.
  • Other tracing applications like Apple and Google’s proposal, DP3T, MIT’s Private-Kit and PACT, Singapore’s TraceTogether generate a new random token to be used as a fresh id after a pre-specified interval.
  • AarogyaSetu also collects more metadata compared to the other apps.
  • Metadata includes details as the timestamp of the contact, the MAC address, the Bluetooth model name and number of the contacted device.
  • Additionally, while other application (except TraceTogether) assume the centralised server to be untrusted AarogyaSetu, on the other hand completely trusts the centralised server.
  • Both, the static id and the collection of additional metadata by the AarogyaSetu app, especially the time stamps and geolocations make it vulnerable to privacy attacks by users.


5.1 Lack of error model

  • A basic engineering principle states that all measurements must be entailed with associated error model clearly specifying the least count and a confidence interval for the measurement.
  • Similarly, for using technology in risk measurement, precise estimates of the rates of false positives and false negatives need to be specified.
  • AarogyaSetu does not specify such rates.
  • Additionally, there are currently no models or principles for estimating the infection risks for both GPS and Bluetooth proximity based estimation.

5.2 Other drawbacks

  • Additionally, Aarogya Setureveals an estimation of “infection risk" within a radius of 10−500m to its users.
  • Given that the stigma and fear of COVID-19 has outgrown the disease itself and there are several reports and incidences of targeting and stigmatising doctors, service staff, as well as members of vulnerable communities for fear of spreading the virus, using a large radius of 10−500m for risk estimation is unwise.
  • Although the source code of a version of the app is now made public, the design details the underlying conceptual principles and server side details are yet not publically available.

5.3 Final Verdict

  • The combined use of GPS colocation and Bluetooth radio proximity for risk estimation of COVID-19 appears to be a leap of faith.
  • The problem is compounded by low smartphone penetration in India.
  • Too many false positives and false negatives may lead to an unbounded noise-to-signal ratio for infection transmission creating confusion and detraction from the main effort by sending administrators and policy-makers on a wild chase.
  • Without a clearly specified protocols and details regarding the central server and in the absence of a regulatory oversight, illegal identification of users and other violations are also possible at the server.


  • Use of an app like ArogyaSetu for estimating risk of infection at the micro-level is not as effective as a local community based manual contact tracing.
  • The manual contact tracing has been applied to much success in Kerala and Dharavi in Mumbai leading to impressive containments.
  • However, the application of contact tracing method is highly restrictive in cases of community transmission, as many instances of spreading will not be caught by it.
  • GPS based geo-location, however, can be effectively used in identifying hotspots at the macro-level.


Public applications like AarogyaSetu must definitely be more transparent in their design and implementation.

Aarogyasetu is an exemplary use of technology to provide social solutions but the fears of inadequate privacy protection and effective risk assessment can not be ignored.

For an app emerging as a foremost scientific and policy response tool in India's fight against COVID-19, AarogyaSetu needs closer introspection.