The Missing Pieces of Contact Tracing
How Web3 has the Mechanisms for a Successful Contact Tracing Application
COVID-19 has changed the world at an astounding pace — from established institutions to startup unicorns to the corner store — but most of all, it has jarringly changed the day to day rhythms of many people. The biggest question right now is how do we continue to contain COVID-19 and prevent future outbreaks of it or other deadly agents. Contact tracing has been the go-to solution to contain outbreaks in this crisis but current methods are lacking many factors for long-term success. At Conflux Network, we see Web3 technology as a solution that solves the shortcomings of current contact tracing solutions.
Shortcomings of Current Solutions
Currently, contact tracing is a manual process where once diagnosed, people are asked to follow-up with people they have come in nearby contact with. This has many flaws: the biggest being that it requires a person to remember who they have been near. In a time where social-distancing is recommended, this may seem more trivial, but as the world re-opens, there needs to be a better method. Another solution currently being implemented in Asia is QR codes where individuals are assigned a QR code that is scanned at any public places they visit. It is a step in the right direction of digitizing the tracking; however, these codes are very directly associated with a person and lead to a big brother-esque feel to the system, and again this requires manual interaction to track an individual contact.
The latest system in development is a collaboration between two technology giants — Apple and Google. They propose to use a privacy-preserving contact tracing system where developers can build applications that integrate with a provided API. In simplest terms, application users would use Bluetooth Low Energy (BLE) technology to exchange identifiers that change every 15 seconds. This identifier is stored locally, the API is used to provide the randomized identifiers of infected individuals that have been reported to the Google/Apple-run database, and then a user can be notified if there are any alerts. While this is a huge leap, there still are concerns about privacy and usability.
Privacy is a huge concern because of the centralization of information on infected individuals — especially with technology giants operating in a data economy that is outside of user influence. Additionally, as global concern over COVID-19 normalizes, what are mechanisms that would incentive continued use of a contact tracing system beyond moral/societal obligation? Web3 has the opportunity to answer those concerns as well as technology that enable contact tracing to become more widely adopted.
Tokenization of Contact
The general idea is to tokenize contact between people; in other words, every time a person comes close enough to another person data is exchanged that represents that instance of contact.
Contact tracing by trading digital information is very similar to Apple and Google’s platform; however, this removes the centralization of information by large corporations.
In a tokenized state, the data is a transaction immutably recorded on the blockchain. Additionally, by tokenizing interactions, other measures can be introduced to improve user retention and decrease privacy concerns. More technical specifications and mechanisms will be discussed in the conceptual build-out.
Conflux Network and our strategic partner Celer could form the basis for a mobile application that tokenizes contact. Conflux Network has many benefits including high TPS, but in this case, transaction sponsorship and the utility value of the CFX token hold more weight.
Transaction sponsorship has the potential to make user interaction even easier because they no longer need to worry about gas fees that typically mark a blockchain transaction. Additionally, users could be incentivized to remain active participants with rewards of CFX tokens. And the power of CFX tokens lies in the growing number of protocols we’re building and integrating.
With those rewards from participating in contact tracing, users have access to a whole ecosystem of financial tools, and even the resources to develop new protocols.
Celer is a layer-2 solution that uses a combination of state channels and sidechains to harness the best of both scaling technologies. The state channel network is built on three layers: cChannel for state channels between two entities, cRoute to form intermediate connections between two non-directly connected entities, and cOS which is the interface to build applications using their web and mobile SDKs. Traditionally, channels only work when both parties are online to provide instant confirmation.
Celer solves this problem by incorporating sidechain technology as their State Guardian Network (SGN) which handles any dispute cases and if one participant in a channel is offline.
Through these two mechanisms, Celer provides even higher transaction rates that can accommodate the large numbers of contact tracing transactions.
While Conflux Network has a transaction rate of close to 5000 TPS, a solution like Celer is still required for the sheer number of transactions required for a contact tracing application to run in real-time. According to Gallup, Americans had a mean of approximately 10 contacts per day between March 27 and April 9, 2020 (contact is defined as coming within 6 feet of a non-household person). If this number is used as a baseline and multiplied with the number of people in the world (~7.8 billion) who are between 15–64 (65%), then converted to contacts per second, the result is ~586K contacts per second. This number is well beyond the capabilities of modern blockchains, but it is also highly unrealistic because it assumes 100% worldwide adoption by adults between 15 and 64. If only 10% of that group adopts the contact tracing technology, it would still be ~59K contacts per second. However, that number does not account for other cultures that have a larger emphasis on social gatherings than America, and the increase in contact as lockdowns across the world are beginning to loosen. Additionally, when contact is made, there are two transactions to send tokens from one to the other and vice versa. Clearly, for real-time contact tracing to work, a layer-2 solution like Celer is needed.
Conceptual Build Out
The following concept is a contact tracing application that has core functionality built on Conflux Network. It does not include many references to Celer because Celer would replicate similar behavior on their state channel and sidechain network to facilitate faster real-time transactions of contact token transfers. The concept uses similar technology to Google and Apple’s implementation: BLE packages sent to nearby devices and the packages are constantly randomized. Then users use the BLE package to check against a “database” to check for risk of infection.
When a user first connects a wallet and registers, a new token contract is created specifically for them. This token contract creates a marker (or token) when the individual interacts with another, and the address of the user and their token contract are stored on-chain. Each time the mobile application changes the user’s BLE payload, the hash of the payload is mapped to the user’s token contract address on the blockchain network as well as stored locally on the device. Each hash of new BLE payloads acts as “sub-wallet” addresses that the user can use to track contact but are not directly connected to them. Additionally, users can report their infection by changing a parameter on their personal token contract.
If the user comes in contact with another individual, BLE payloads will be swapped and stored locally in the application. The received BLE payload is then hashed, and a token is sent to the generated address. When the application is checking contacts, it can use the hash of the received BLE payloads, and send it to the registration contract where the BLE hash address is internally mapped to a token contract address. The registration contract can use the respective token contract address to check the status of the contract and return it to the application. If the infection parameter is true, the application can check the user’s token contract for the tokens sent to the BLE hash address. This returns token transaction timestamps and quantities which can be synthesized to determine the risk level.
The two major improvements are a reward mechanism and privacy.
The reward mechanism creates a monetary incentive for users to continue to use the application, and the CFX reward has utility on a much larger platform — it can be spent, converted, staked, or used for development. The funding method for the rewards could be staking interest on Conflux Network, outside investors, or other fundraising methods.
Compared to traditional mobile apps that utilize the Google/Apple API methods and offset costs through in-app advertisements or selling data, this is a massive improvement.
Privacy is improved by using the randomly generated BLE payloads and on-chain mechanisms. By sending tokens to a sub-wallet, a user cannot know the address of the person they sent the token. Additionally, the registration contract can have all of the mappings as private. It would be storing two mappings: user address + token address, and BLE hash address + token address. By keeping these private, the only way to monitor the system is to watch transactions on the registration contract. While not a perfect system (it can still be defeated by large-scale data mining), there are still many privacy improvements that could be made such as zero-knowledge algorithms to mask transaction parameters. Additionally, by using Celer’s state-channel and sidechain technology, there can be further measures of privacy by hiding intermediate transactions and only reporting final states to Conflux Network.
Room For Improvement
There is always room for improvement and further development, especially in a conceptual application. The biggest improvement in functionality would be integration with existing systems like Google and Apple’s API. This is possible as long as the BLE payloads follow the same standard and a method of reporting infections to the blockchain is implemented. Additionally, privacy algorithms are a constantly evolving space and are an area of opportunity for huge improvements, especially in contact tracing applications. And finally, how to bring contact tracing and Web3 connectivity to non-smartphone devices to ensure that there is equal access to tracing technology for any demographic.
At Conflux Network, we recognize that the world is changing at an incredible pace as we endure and fight the COVID-19 pandemic. And as the world comes to terms with a new reality that includes social distancing, masks, and contact tracing, we believe that it does not mean an end to privacy. Instead, this is an opportunity to use Web3 and blockchain technology to build on the lessons learned from Web2 and blockchain predecessors to preserve privacy now and for the future.
Written by Conflux Network’s Research Engineer Aaron Lu
Celer Network is a leading layer-2 scaling platform that enables fast, easy, and secure off-chain transactions for not only payment transactions, but also generalized off-chain smart contracts. It allows anyone to quickly build, operate, and use highly scalable decentralized applications through innovations in off-chain scaling techniques and incentive-aligned cryptoeconomics. For more information, visit: https://www.celer.network/, and join the official Celer Network community on Telegram: http://t.me/celernetwork.
About Conflux Network
The most endorsed DLT project in China, Conflux Network is an open-source, layer-1 blockchain protocol delivering heightened scalability, security, and extensibility for the next generation of open commerce, decentralized applications, financial enterprises, and Web 3.0. Conflux Network is overseen by a global team of world-class engineers and innovative computer scientists. Fostering entrepreneurship and innovation, Conflux unites startups and enterprises across industries and continents to promote decentralized collaboration in advancing blockchain technologies for real-world solutions. Founded in 2018, Conflux has raised $35 million in capital from prominent investors including Sequoia China, Metastable, Baidu Ventures, F2Pool, Huobi and IMO Ventures.