Study of Blockchain based Digital Platform for Offering Public Distribution System Services through General Trade Outlets and Direct-to-Home delivery in India
Introduction
In our first post we, AsterQuanta [AQ], introduced the public distribution system [PDS] in India and how we obtained permissions from the government for enabling proxy distribution of ration services.
In our second post we presented a survey that captures the time, effort and cost aspects from a beneficiary perspective and insights from the pre-deployment solution development for proxy distribution of rations.
In our third post we described the distribution models, operational stack and processes, findings and key learnings from testing the distribution models in the field while delivering the technology & design for enabling proxy distribution.
Figure-1 captures the current and proposed process of food grains distribution. In the current system, food grains are sourced from Food Corporation of India (FCI) warehouses by the FPSs to be distributed to the beneficiaries. This is shown in the upper part of the picture. Multiple FPS outlets are directly stocked with food grains by FCI. Subsequently beneficiaries access these food grains from any of these FPS outlets.
In our pilot, Village Level Entrepreneurs (VLEs) sourced food grains from these FPS outlets and directly delivered to the beneficiaries or through the retail outlets. For our pilot, studies and experiments, we have designed an operations stack with a virtual FPS engine that can sub-source from the FPSs (accessed by the VLEs) and then distribute it to retail outlets for further decentralization or disburse directly to the beneficiaries enabling Direct-to-Home delivery. The neighborhood retail outlets also worked as intermediate distribution nodes for the beneficiaries simplifying the access to the rations.
Issues
Misuse of entrusted power for private gain[1] is a challenge in public aid distribution scenarios. Various central and state governments in India have been putting efforts and resources in trying to address this challenge.
Distribution channels are entrusted with the power to source and disburse aid items. Studies indicate that when the distribution channels have local monopolies i.e. there are none or limited options for beneficiaries to avail their aid from, leakages become rampant.
Similarly in the PDS scenario, with only one ration aka fair price shop in a local area, beneficiaries not only lose out due to leakages , but also have to endure the vagaries of the PDS owners & associates.
We were able to identify potential leakage points in the current process of food grains distribution. The primary leakage points are at the FPS distribution level. They are –
a) Under weighing of grains and items
b) Not all beneficiaries get the entitled benefits
There is a potential chance of loss during transportation from warehouse to the fair-price shops. The issues are highlighted in Fig.2
Addressing the Leakage Issues
Currently government has enforced digital mechanisms (digital weighing machines, Biometrics etc) to contain the above leakages. But they are not completely foolproof as these are open-loop measures with no adequate controls in place to check their compliances. For example- the digital weighing machine can be manipulated to over-weigh grains and items put on the machine. There is no digital reconciliation between what is being distributed using the weighing machine and what is arriving at the fair price shop.
● To address these challenges, our pilot project had been planned around the use of blockchain technology to make disbursal of public goods fair and secure
● Early on in the process, we realized that a live implementation of blockchain technology would rely on access to government held data
● In order to test our idea, we devised a system working around this requirement
● In this article, we explain why integration of blockchain tech will still be crucial to lasting change in the PDS system, and why we believe that through our work to date we have investigated and created the conditions necessary to do so successfully
Blockchain in aid distribution and public distribution systems (PDS)
A blockchain is a distributed, or decentralized, ledger — a digital system for recording transactions among multiple parties in a verifiable, tamperproof way. The ledger itself can also be programmed to trigger transactions automatically. Cryptocurrencies that have been designed to replace fiat currencies work on blockchain technologies.
Blockchain helps in making third party entities, systems, or personnel that enable contract execution through manual processes become redundant i.e. eliminating human interventions. Blockchain technology manifests into smart contracts. Smart contracts are self-executing contracts which contain terms and conditions of an agreement between parties and/or peers. Smart contracts
+ are written in code
+ are executed in block chain’s infrastructure
+ have multiple execution points agreeing on a common result, thereby making it difficult to manipulate.
+ have results recorded in a distributed fashion i.e. distributed ledger
+ are immutable. So they will not let any person or system alter the agreement
Globally, aid distribution faces leakages, corruption & distribution bias challenges. Many processes, solutions and systems have been implemented to address these challenges. These solutions are primarily implemented in an open-loop fashion. i.e. feedback from end beneficiaries are not used as a factor to address the challenges. Feedback from the end beneficiaries and users in the form of aid utilization information and/or aid utilization events will greatly help in inventory and financial regulation.
Utilization at the beneficiary level can be sensed using multiple ways.
Some of which are –
a) Health indices
b) Social living indices
c) Beneficiary behavior
d) Sustainable Development Goals (SDGs)
e) Other social inputs
We focus on aggregated beneficiary behavior in a decentralized setting to sense utilization events & utilization information.
Fig-3 showcases the fundamental difference(s) between the current open loop controls for addressing leakages/challenges and closed loop regulation based on utilization at the beneficiary level.
Figure-3: Delivery to Beneficiaries —
Decentralization of distribution at the last mile is key to generate multiple flows of information (with ration-aid, obviously) which can be utilized for reducing leakages. This information has to be captured in a fool-proof manner.
Since information flow, inventory flow and financial flows can be captured in an automated fashion using blockchain and self-executing contracts, we propose the same to address ration distribution (PDS) leakages and other challenges.
Last mile decentralized public distribution system (PDS) (aid distribution)
Transforming available frontier technologies into value added business (aid distribution) transactions is a challenge. We are highlighting how through our studies, implementation and proposal, we are addressing this challenge.
As described in the previous section, multiple distributions channels (aka decentralized distribution) leads to multiple flows of information. Without multiple channels, there is no variance in information that can be utilized for optimizing the distribution. i.e. with just one channel of distribution, the feedback points and control mechanisms are very limited and there is not much to optimize (challenging to reduce leakages and corruption).
In our pilot, we designed and deployed a decentralized last mile distribution. This was key to explore blockchain based ration distribution. Decentralization was achieved by creating two parallel channels of distribution — through retail outlets and direct-to-home delivery. Without these two additional channels of ration distribution, just block chaining the transactions in a single distribution channel setup does not serve much purpose as the block chained transactions do not carry information about the end utilization of the ration.
Village level entrepreneurs (VLEs’) sourced food grains from FPS outlets and directly delivered to the beneficiaries or through intermediate retail outlets. For our pilot, studies and experiments, we have designed an operations stack with a virtual FPS engine that can sub-source from the FPSs’ (accessed by the VLEs’) and then distribute it to retail outlets for further decentralization or disburse directly to the beneficiaries enabling Direct-to-Home delivery. The neighborhood retail outlets also worked as intermediate distribution nodes for the beneficiaries simplifying the access to the rations.
In the above mentioned decentralized distribution setting, we propose the following blockchain based environment for reducing leakages and address other challenges.
Block Chain Smart Contract Execution
⮚ L1: When pull thresholds are met (demand from beneficiaries)
⮚ L2, L3: Programs on block chain automates execution of supply chain from FCI to retail/outlet/distributor
⮚ Smart contracts are self-executing contracts which contains the terms and conditions of an agreement between the peers
Leakage reduction through
⮚ Smart contracts are self-executing contracts which contains the terms and conditions of an agreement between the peers
⮚ Terms and conditions are written in code
⮚ Code is distributed and executed in BC’s infrastructure
⮚ Result is recorded on distributed ledger
⮚ Smart contracts are immutable, hence no individual can alter the agreement
Simulation:
Red retail icon indicates a distribution channel that is unfair with lesser number of people utilizing it for their benefits
Green retail icon indicates a distribution channel that is fair with more number of people utilizing it for their benefits
Algorithms learn to reward the “fair distribution channel” with more opportunities to distribute compared to the “unfair distribution channel”. i.e. more ration is sent to the green icon for distributing the ration to the larger number of people that visit this particular outlet.
Technology Stack: Block chain and Artificial Intelligence (Reinforcement Learning)
The last mile distribution setups contain a lot of dynamism. Dynamism can manifest in various forms such as people’s behavior changing with respect to accessing the aid, addition of new channels or deletion of old channels that are used for sourcing aid, fair channels becoming unscrupulous over a period of time due to monopolistic situations, cultural and political changes etc.
There is a need for automated mechanisms to address the dynamic environments in which the distribution happens. The mechanisms have to be self-learning in nature and have to work sometimes in an exploratory mode, while rewarding the fair distribution channels and penalizing the unfair ones. Artificial intelligence algorithms such as reinforcement learning are well suited to handle such dynamic environments and can help in optimized distribution of the aid/ration.
What next?
We hypothesize that an ecosystem with the following characteristics would provide substantial benefits in leakage reduction
a) Last mile distribution not inextricably linked to decisions made by central authorities
b) Consensus mechanisms at the local level to regulate aid distribution
The challenges to creating such an ecosystem would be from
a) resistance from local ration shops
b) digital literacy and learning requirement for individuals/entities to become channels
c) fine tuning operations to handle diverse dynamics across geographies & cultures
d) potential cost increases, and
e) who bears the cost?
Testing the hypothesis
Our initial work was focused on surveying and collecting information from beneficiaries. We stimulated the beneficiaries’ thought processes by presenting possible opportunities.
Subsequently we worked on their responses and selected 100 beneficiaries for a 3 –month pilot. Through real and on-the-ground distribution channels (direct-to-home distribution & retailer distribution), we collected feedback from these beneficiaries. One of the stress-test questions we posed to beneficiaries was whether they would be willing to pay for these services (direct-to-home and retailer channels were provided as a free service during the pilot).
70%+ beneficiaries reported willingness to pay for these newly created distribution channels. Ration was free. But people were willing to pay extra just for more access options. I.e. We postulate that the willingness to pay as evidence of the value beneficiaries place on choice & convenience. Hence, we also postulate that by providing people more choices to access their ration/aid, an invisible hand (consensus) is created. This invisible hand allows the market to reach equilibrium without government or other interventions forcing it into unnatural and manipulated patterns [4].
As a next step, we would like to deploy an integrated AI-Blockchain system in place and test the system on 1000+ beneficiaries.
Conclusion
Aid leakage (i.e. quantity of ration distribution leaks) can reduce by
a) Providing consumers more choice by decentralizing aid (ration) distribution channels, especially at the last mile.
b) Empowering consumers to choose from multiple options. Multiple options (distribution channel) are created by reducing constraints and requirements for an individual or a channel to distribute. Aka eliminating unintended government granted monopolies.
c) Allowing consumer choice to regulate aid distribution; more supply sent to outlets with greatest demand
d) Using Blockchain & AI to optimize distribution from fair price shops to last mile outlets.
Bibliography
[2] Harvard Business review report on Operations and Supply Chain management — Building a Transparent Supply Chain by Vishal Gaur and Abhinav Gaiha.
[3] A Smart Card Solution for India’s Public Distribution System (PDS) — gesellschaft für internationale zusammenarbeit (giz) gmbh
[4] https://www.investopedia.com/terms/i/invisiblehand.asp
[5] https://www.gemini.com/cryptopedia/fiat-vs-crypto-digital-currencies#section-issuance-and-governance
[6] https://www.nielsjohannesen.net/wp-content/uploads/AJR2020-revised.pdf
