This article is part of The Wire‘s ‘India Black Boxed’ series. Read it here: Introduction | Part I | Part II | Part III | Part IV | Part V
In parts one and two of this series, I sought to probe deeper into several answers which the Election Commission of India (ECI) has provided in its revised set of 100 frequently asked questions to demonstrate how unsatisfactory they are.
The ECI publicised these revised FAQs on its website in January 2024 with the objective of clarifying all doubts about these electoral machines.
When compared with the information contained in publicly available official documents, the ECI’s claims about the design, and functioning of electronic voting machines (EVMs) and Voter Verifiable Paper Audit Trail (VVPAT) units raise more suspicion instead of resolving persistent doubts.
In this write-up, I address a few more of ECI’s claims in light of patent-related documents and the information disclosed in response to formal requests for information made under the Right to Information Act, 2005 (RTI Act) with particular focus on the following grey areas in the FAQs:
- Why are the ECI’s Technical Expert Committee members also listed as VVPAT inventors?
- How transparent is the independent review of EVM-VVPATs conducted by the Standardisation Testing and Quality Certification?
Why are ECI’s Technical Expert Committee members also listed as VVPAT inventors?
The ECI has claimed that the EVM-VVPAT combo is manufactured by two Central Public Sector Undertakings (CPSUs), under the guidance of Technical Experts Committee which it has constituted.
In response to FAQ #7: “Where are the EVMs manufactured? Are they imported?” the ECI has stated as follows:
“Ans. EVMs/VVPATs are not imported but indigenously designed and manufactured by two Public Sector Undertakings (PSUs) namely Bharat Electronics Limited (BEL), under the Ministry of Defence and Electronics Corporation of India Limited (ECIL), under the Department of Atomic Energy under the guidance of the Technical Experts Committee (TEC) constituted by the Election Commission of India.”
Further, at FAQ #68, the ECI admirably rises to the defence of the competence and credibility of the TEC members as follows:
FAQ 68. ““Experts” have declared that the ECI machines are non tamperable and safe but this does not make the EVM+ [sic] None of the ECI experts have credentials in computer security and the majority of them are not even computer scientists. In addition to experts the ECI is also dependent on many other entities and organisations-including hardware manufacturers, software developers and testers, system assemblers, and un-modelled custody chains for the integrity, safety and security of its machines and is thus not entirely in control. In this situation what is the need to use EVMs for elections in India?
Ans. Comments on ECI experts are uncalled for. The EVM detractors, self-appointed technical experts and certain social media personnel have repeatedly failed to realise that the EVMs cannot be compared to a computer which runs on an operating system. The microcontrollers of EVMs and VVPATs run on a specific program designed to faithfully record the voters’ choice.
That the EVM is a truthful machine has been proved over decades of usage through changes in several governments at the state and central level, whenever the public wanted to. In fact, the EVM has made elections safe and virtually eradicated booth capturing by limiting the rate of vote casting to four votes a minute and thus significantly increasing the time required for stuffing false votes. Invalid votes a bane of the paper ballot system was at times greater than the winning margins and have been completely eliminated by the EVMs.
Vote cast is verifiable by voter on VVPAT and the Control Unit-VVPAT vote count matches are done on well-established principles of statistics to provide a very high level of confidence.”
In its hurry to dispel all doubts about EVMs and VVPATs, the ECI has left the first sentence of FAQ #68 hanging, but let me not quibble about this minor issue. In response to FAQ #1: “What is an EVM”, the ECI hyperlinks a portion of its answer to a page in its 2021 Status Paper on EVMs and VVPATs containing the names of the TEC members and their appointed role (see Images 1&2 below):
Images 1 and 2: Pages hyperlinked to FAQ #1 explaining the membership and role of ECI’s TEC.
Originally constituted in 1990, the membership changed about 15 years later. The current TEC was constituted in 2010 with the following experts:
- Professor D.T. Shahani of the Indian Institute of Technology, Delhi;
- Professor Rajat Moona of the Indian Institute of Technology, Bhilai;
- Professor Dinesh Sharma of the Indian Institute of Technology, Bombay; and
- Professor A.K. Agarwala of the Indian Institute of Technology, Delhi.
I will not question their scholarship or competence as I am not qualified to do so. However, as a voting and taxpaying citizen of India I have every right to ask this fundamental question.
Why are these current TEC Members also included in the list of ‘inventors’ of VVPAT?
Images 3 and 4: Declaration of ‘Inventorship’ submitted by BEL to obtain a patent for the VVPAT.
In the Inventorship Declaration Form filed with the Patents Controller while applying for a patent, BEL listed 12 individuals as the “Inventors” of VVPAT.
All four TEC Members are also listed as Inventors of the VVPAT.
Further, according to its Proof-of-Right document submitted to the Patents Controller along with its application, BEL calls itself only the “Applicant” for the VVPAT patent but not its “Inventor” (see Image 5 below). All 12 individuals listed above are shown as “Inventors” here as well.
The names of all four TEC Members are found in this list (see Images 6-7 below). This document also bears the signature of one of the TEC Members (see Image 7 below).
The role of the TEC is clearly specified in ECI’s 2021 EVM Status Paper as follows (see Image 2 above):
i. Give technical advice to build specifications and design of newer versions of EVMs/VVPATs that they incorporate latest technology both in hardware and software design and improving robustness against tampering.
ii. Examine design proposals of manufacturers of EVMs and offer recommendations for improvement.
iii. Mentor designs process wherever asked.
iv. Examine concerns raised on EVMs tamperability.
v. Any other advice that Commission [ECI] may seek or any other technical work that the Commission may entrust from time to time.”
Further in response to FAQ #33 (see Section II below) the ECI states that the TEC also conducts an independent audit of the voting systems design.
Can TEC Members sit in judgement over their own invention? Is this not a huge conflict of interests? Is it not reasonable to expect that inventors will defend their invention against all objections raised by anybody and everybody?
In fact, in December 2020, the Patent Examiner submitted a report objecting to BEL’s claims that the VVPAT was its unique invention and completely dissimilar to other electronic voting and vote counting machines being used elsewhere on the planet.
Images 5, 6 and 7. Pages 1-3 of the Proof-of-Right document BEL submitted along with the VVPAT patent application.
The Patent Examiner also remarked that the only innovative thing about the VVPAT claim was its software but under Section 3(k) of The Patents Act, 1970, software is not patentable. BEL sought extra time to make its submissions in response to these objections and eventually obtained the patent in May 2022. The Patent Examiner’s report and BEL’s submissions are all available on the Union government’s website.
Did the TEC Members give any advice to overcome the Patent Examiner’s objection? The patent records by their very nature will not reveal such details. Such information will be available in the records held by BEL and the ECI. However, neither of them has disclosed any details about the role that TEC Members played in overcoming the Patent Examiner’s objections. In fact, none of the TEC reports with regard to VVPATs are available in the public domain.
In response to my RTI intervention of March 2019, the EC furnished TEC reports pertaining to EVMs only, prepared between 1990 and 2013 and no further. This is another instance of the ECI and BEL putting away crucial technical information including TEC recommendations about VVPATs in a black box. The FAQs are simple inadequate for arriving at an informed opinion about the EVM-VVPAT combo.
Now how is this different from the patent that BEL and ECIL obtained in 2006 for their EVMs? The name of BEL’s inventor, Rajagopalan Jagannathan, is mentioned in the patent summary on the IP India website (see Image 8 below). ECIL also mentions only one individual, Gadde Raja Koteswara Rao, an employee, as the inventor of its EVM (see Image 9 below). Neither individual ever served on any of the earlier TECs (see Image 1 above).
The TEC’s role at that point of time was limited to providing technical advice and there was no occasion for any conflict of interests to arise. However, there is a clear case of conflict of interests with the current TEC members vis-à-vis their role as co-Inventors of the VVPAT. Strangely, the Election Commission does not talk about any of these issues in its FAQs.
Images 8 and 9. Patent documents showing the names of the EVM inventors of BEL and ECIL.
How transparent is the independent review of EVM-VVPATs conducted by STQC?
The ECI explains the systems put in place for an independent review of the functioning of the EVM and VVPAT combo in response to FAQs #33 and #72 as follows:
“Q 33. Can the ECI implement an independent review system to review the voting system and the integrity of election process?
Ans. The voting system designs go through an independent review and even an independent audit by Technical Expert Committee. Various design details are also available on the site at a level that the stakeholders may be interested in. The outcomes accepted by voters and candidates are the biggest form of review (Annexure-II and Annexure-III).
The ECI-EVM voting system goes through an independent review by way of third party (STQC) checks. They also go through independent audit by various political parties/candidates/their representatives like in mock polls. The processes are well defined and are put on the website of the ECI for public view.”
“Q 72. While banning electronic voting, the German Constitutional Court made the following observation: The use of voting machines which electronically record the voters’ votes and electronically ascertain the election result only meets the constitutional requirements if the essential steps of the voting and of the ascertainment of the results can be examined reliably and without any specialist knowledge of the subject… The legislature is not prevented from using electronic voting machines in elections if the possibility of a reliable examination of correctness, which is constitutionally prescribed, is safeguarded. A complementary examination by the voter, by the electoral bodies or the general public is possible for example with electronic voting machines in which the votes are recorded in another way beside electronic storage. How are ECI EVMs different from the voting systems banned by the German Constitutional Court?
Ans. Across the world, both paper ballot voting system as well as EVMs are used for conducting elections as per the preference and mandate of the authorities concerned. The Election Commission of India is a Constitutional Body mandated to conduct elections to State Legislature, both the houses of the Parliament and the offices of the President of India and Vice-President of India, as per Act and Rules passed by the Parliament of India. Use of ECI-EVMs for conducting elections in India is approved by the Parliament and upheld by various High Courts and Hon’ble Supreme Court of India.
ECI EVMs are manufactured by central government public sector undertakings in a secure manufacturing facility. Rigorous third-party testing is carried out by STQC (Standardization Testing and Quality Certification) at the manufacturing premises before acceptance and dispatch to various states of deployment. Stringent and elaborate protocols as mandated by ECI are followed during EVM movement, storage, and deployment. The German Constitutional Court made its observation in the context of EVMs used in German elections and in relation to German law. Indian EVMs are robust and implement technologies and processes which are different and noncomparable. Hon’ble Supreme Court of India and various High Courts have time and again scrutinised the machines and have reposed their confidence and faith in ECI EVMs.”
The sum and substance of the ECI’s responses to the FAQs above is that STQC is the agency that is officially tasked with the review and testing of EVMs and VVPATs. STQC (Standardisation Testing and Quality Certification) is a Directorate attached to the Union Ministry of Electronics and Information Technology which provides quality assurance services in the are of Electronics and IT through a country-wide network of laboratories and centres.
The STQC claims that all of its financial obligations are met through budgetary support provided by the government of India, so it is essentially an agency of the Union government. The STQC also claims that ECI has identified its Electronics Test and Development Centre (ETDC) Hyderabad as the testing lab for third party testing of EVMs and VVPATs. However, no further information about what kinds of testing STQC does is available on its website.
In June 2019, an RTI application was submitted to the STQC seeking information about the names and designations of the STQC team which conducted testing and evaluation of the firmware of the EVMs and VVPATs used during the 2019 Lok Sabha elections; the dates on which such testing was carried out and the geographical location where such testing was done; the total number of BEL and ECIL manufactured machines which the STQC team tested and evaluated, a copy of all manuals and standard operating procedures (SOPs) which contains the methodology to be used for such testing purposes; and a clear description of the instruments used for such testing, a copy of the test reports prepared by the STQC team.
Data about the number of EVM-VVPATs which showed mismatch with the reference firmware and other kinds of errors detected during the testing was also sought.
The STQC rejected all these queries invoking Section 8(1)(d) of the RTI Act which protects information that is in the nature of commercial confidence or trade secrets or intellectual property rights (IPR) whose disclosure would harm the competitive position of a third party (see Images 10 and 11 below).
Images 10 and 11: STQC’s reply to RTI queries about EVM and VVPAT third party testing.
When this reply was challenged before the first appellate authority of STQC, he upheld the denial on grounds of commercial confidence without specifying whose interests were being protected – STQC’s, ECI’s, or those of BEL and ECIL. So, this matter was escalated to the Central Information Commission (CIC) the same year.
The CIC took two years to hear this case. During the hearing STQC opposed the second appeal claiming that testing services are provided by levying charges on the requestor for such services, so the information sought in the RTI application belongs to the service requestor. Disclosure would harm the competitive position of BEL and ECIL, STQC argued (see Image 12 below).
Image 12: STQC’s counter to the second appeal submitted to the CIC.
Among the several counter arguments that I submitted to the CIC, I had contended that there is no possibility of any harm arising on account of the information disclosure because the two CPSUs have a ECI-guaranteed duopoly on the EVM-VVPAT market in India. In July 2021, the CIC directed that information regarding testing dates and quantities of machines tested be disclosed as no harm would be caused. It refused to direct disclosure of the names and designations of the members of STQC’s testing team holding that it is in the nature of commercial confidence. Instead of deciding on their correctness despite having the power to do so, the CIC merely directed STQC to re-examine its replies to other queries.
In August 2021 STQC sent a revised reply (see Images 13 and 14 below) providing access only to the period during which the machines were tested at the factory premises of BEL and ECIL. Despite the CIC’s finding that statistical information about the number of EVMs and VVPATs tested does not attract any exemption, STQC rejected this information under Section 8(1)(e) now, claiming fiduciary relationship with BEL and ECIL. It also rejected access to manuals and SOPs used for testing purposes.
The STQC said that no errors or mismatch were found in the firmware used in the EVM-VVPAT combo when compared with the reference firmware. It did not provide a direct reply to the query about other kinds of errors in the machines that might have been detected. Instead, the STQC said that no failures were observed related to the firmware of EVMs and VVPATs it tested.
Images 13 and 14: STQC’s revised RTI reply after the CIC remanded the case back.
Why is STQC’s revised RTI reply problematic?
First, STQC transgressed the CIC’s order by withholding information about the number of machines it tested even though the CIC directed its disclosure. But let us leave that aside for the time being.
Second, STQC provided only the period during which it had tested the VVPATs of both companies. It remained silent about tests conducted on Control Units (CUs) and Ballot Units (BUs) even though similar information was sought about the testing of EVMs. Why is this problematic? This is because of the large number of EVMs and VVPATs that were found defective prior to and during the 2019 Lok Sabha elections.
ECI’s FAQ #99 deals with this issue as follows:
“Q 99. It is claimed that higher percentage of EVMs and VVPATs got defective during First Level Checking of EVMs and VVPATs pertaining to Lok Sabha 2019. What action taken by Election Commission of India to reduce the percentage of defects of EVMs and VVPATs?
Ans: In General Elections to Lok Sabha and simultaneous elections in the States of Andhra Pradesh, Arunachal Pradesh, Sikkim and Odisha, 37,377 Ballot Units (1.8%), 57,775 Control Units (3.8%) and 85,905 VVPATs (5.4%) were rejected during First Level Checking.
However, during actual poll in Lok Sabha 2019, replacement rates were BU – 0.74%, CU – 0.79% and VVPAT – 3.36%.
After every election, an analysis of non-functional EVMs and VVPATs is carried out as an integral part for enhancing the performance of EVMs and VVPATs to reduce replacement rates during actual poll.
Post General Election-2019, the Technical Expert Committee (TEC) carried out an exhaustive analysis along with Bharat Electronics Ltd (BEL) and Electronics Corporation of India Ltd (ECIL). Analysis got delayed due to COVID-19 pandemic period. In analysis, it was concluded that some corrective actions are required to be taken to reduce the replacement rate of M3 VVPATs and to ensure voter is not put to any inconvenience. (emphasis added)
M2 model machines, used during General Election-2019, had been discontinued as their economic life completed. M2-M3 VVPATs, used along with M2 EVMs during General Election-2019, are upgraded to M3 VVPATs, as originally planned to ensure that they are compatible with M3 EVMs.”
ECI’s response to the above FAQ sounds very bureaucratic. Tens of thousands of machines – BUs, CUs and VVPATs which STQC said it had checked between 2017 and 2019, did not pass the first level check (FLC) right up to the eleventh hour when Phase 1 of the Lok Sabha polls began mid-March in 2019. There were more replacements of these machines during the actual polls conducted in seven phases.
In its press release dated March 10, 2019, (see Image 15 below) the ECI published the total number of BUs, CUs and VVPATs that were to be deployed for the 2019 elections (along with general elections to four States and bye-elections to 34 vacant Vidhan Sabha seats in other states and UTs).
Image 15: Extract from ECI’s Press Note dated March 10, 2019.
According to this data, the number of machines replaced during the polls works out as follows:
| Unit Type | Total no. of units deployed | % replaced | Units replaced during actual polls |
| BUs | 23.30 lakhs | 0.74% | 17,242 |
| CUs | 16.35 lakhs | 0.79% | 12,916 |
| VVPATs | 17.4 lakhs | 3.36% | 58,464 |
The data tabulated above starkly reveals thousands more BUs, CUs and VVPATs which had first passed muster during the testing that STQC conducted and next during the first level check which the engineers of BEL and ECIL conducted prior to the commencement of elections, nevertheless had to be replaced during the actual polls.
As voters and taxpayers, we must ask ourselves, if this were the track record of companies selling electronic calculators in the market, would consumers find those machines reliable enough to buy? Would consumers like to spend their money on calculators that are likely to have even a less than 1% probability of being defective or malfunctioning? No, we would want every calculator we buy to work perfectly. Why adopt a different standard for EVMs and VVPATs? This is the elephant in the ECI’s room filled with 100 FAQs which it simply does not address.
I will demonstrate the depth and breadth of this defective/malfunctioning problem in the next instalment of this series.
What kinds of tests did STQC conduct on ECIL’s EVM/VVPAT combo?
In two identical RTI applications submitted to BEL and ECIL submitted in March 2019, I had sought a copy of the reports of the tests conducted on the EVMs and VVPATs in addition to other information. While BEL rejected this RTI query by invoking commercial confidence, ECIL rejected it on grounds of national security. Both matters were escalated to the CIC.
Two years later, the CIC remanded the matter back to the CPSUs to decide the RTI applications afresh even though it has the power to examine the correctness of their replies. BEL reiterated the rejection, but ECIL had a change of heart and supplied a sample copy of the test report conducted by STQC conducted on one batch of its EVM’s CUs (see Images 16 and 17 below):
Images 16&17: Copy of STQC’s test report conducted on the CUs of ECIL-manufactured EVMs.
The test report furnished by ECIL shows only a burn-in test conducted on the PCB (printed circuit board) of the EVM’s CU to review its ability to withstand temperatures up to 55°C. Why did ECIL not disclose the firmware tests that are required to be conducted on the CUs and also the VVPATs is a mystery.
If the hardware temperature stability test report can be furnished under RTI, why can the firmware test report not be shared as well? Why did so many thousands of CUs fail to perform before and during the Lok Sabha polls despite passing the STQC’s tests? The ECI has much more to answer because crucial information about the manner of tests conducted on the EVM-VVPAT combo remain hidden in a black box.
I began part one of this series by recounting the Supreme Court’s exasperation at the number of attempts being made by various petitioners to seek an urgent hearing with regard to the manner in which the 2019 Lok Sabha elections were conducted with the help of these machines. The apex court’s Bench apparently remarked, “Is there no end to these suspicions?”
All that I can humbly say after five years of multiple RTI interventions that there is simply not enough information in the public domain to end these suspicions. This is because the ECI and the EVM-VVPAT manufacturing companies have not adequately discharged their burden of proving their claims about these machines by putting in the public domain crucial official documents. The ECI’s clarifications in its revised FAQs hint at the existence of a whole lot of information that is unfortunately black boxed.
Venkatesh Nayak is Director, Commonwealth Human Rights Initiative, New Delhi. All facts narrated above are in the public domain. Views expressed are personal.
