keyko-io / odyssey-dashboard

Odyssey Air Cargo Flow Challenge - Supply Chain Dashboard

Odyssey Provenance Dashboard

A unique and trusted source of truth for supply chain

Air Cargo Dashboard

Join the Odyssey Hackthon conversation in Nevermined Discord server (#odyssey)

---

• Odyssey Provenance Dashboard
  • Introduction
    • Main Problems
    • What do we need to do to improve that?
    • Benefits
  • Odyssey Demo
    • MSD demo
    • DHL demo
    • KLM demo
    • Final recipient
    • Customs Agent
  • Bad Handover scenarios
    • Scenario 1 - Failed due diligence
  • Dev quickstart
  • Nevermined integration
    • Services
    • Accounts
    • Assets
  • Links

---

Introduction

Fake and counterfeit medications kill tens of thousands of people around the world each year, with profits funding other criminal activities.

While this has always been a hot topic, the COVID-19 vaccine has really put it in the spotlight: imagine the first batches of vaccines arriving after being compromised, tampered with or not kept cool throughout their journey. How would that affect global trust in a COVID-19 vaccine?

COVID-19 has revealed the crucial role of the supply-chain within the global economy. In addition to security, speed and flexibility are now more critical than ever.

Within the context of COVID-19, effective data sharing and transparency are the way forward to improve collaboration within the supply chain. If you want to be fast while mitigating risks, you need to be a team-player. You cannot afford to wait a week to release your data, even towards a competitor.

Main Problems

Some of the main problems identified in a typical air cargo flow are:

• There is no unique source of truth. There are multiple isolated and unrelated systems everywhere
• There is no complete provenance record that can be trusted
• The product associated can be tampered or mixed with counterfeits during transportation
• The customs office needs to know upfront that some goods are going to need to be validated
• Not much transparency. If something goes wrong, it’s difficult to understand how this happened
• It is difficult to recognize a bad actor and mitigate the impact of their actions
• There is no an effective and secure data sharing between parties
• Physical world is badly connected to the digital world

What do we need to do to improve that?

During Odyssey we worked in an on-chain Provenance registry based in the W3C Provenance specifications. The W3C worked for providing a generic and standard able to track the information about entities, activities and people involved producing data or things.

Our motivation here is to provide a unique source of truth for all the parties involved during the cargo process. This is gonna be based in a Smart Contract and will include:

• The complete provenance record during the whole cargo life-cycle. This will provide a complete lineage of the goods during the cargo flow, since the manufacturing until the final delivery.
• The digital signatures of all the parties involved. To avoid the goods tampering, during the handover of the goods between parties the signatures of the parties involved are collected. In the same provenance record the solution is gonna keep these digital signatures identifying the entities involved.
• The integrity fingerprint of the digital assets involved. Any physical good going through the cargo flow has associated multiple digital files and metadata. This can include manufacturing details, product specifications, quality controls, etc. We are gonna record the fingerprint of these digital assets to flag if any of them are changed afterward.

Benefits

What are the main benefits of it?

• In this solution, digital signatures are required for each step in the supply chain process. It makes possible to identify if an unexpected actor is taking the cargo and at what point happened.
• We make available an on-chain Provenance record as a unique source of truth. This provides a complete lineage of the cargo.
• Existing and independent Supply Chain Management (SCM) software components used by the different entities participating in the supply chain process, can be plugged to the network without modifications, and make information available to specific users.
• Notifying the customs office upfront, we can facilitate the process of goods going through it
• Transparency, the complete transactions record is kept in unique and decentralized place where everybody involved (having the right permissions) can get access to relevant information generated during the supply chain process
• It allows an automatic way of sharing digital assets and metadata related with the cargo (like quality control checks, receipts, customs clearance documents, etc.)
• Analytics and federated learning across all the parties data

Odyssey Demo

During the Odyssey demo we want show how the multiple parties participating in the air cargo flow can use a common backbone to track and manage the provenance record of the cargo. This brings the benefits described above.

MSD demo

The flow of the demo using a MSD user is:

1. We open the application as a MSD user
2. We show the MSD home page with all the goods managed by MSD and the different states (Registered, In transit, Delivered, Problem detected)
3. We click the '+' to register a new good
4. The app opens a screen with the camara enabled, the MSD user scans the good QR code
5. After the QR is scanned, a screen asking for some additional metadata (name, description, etc.) is shown
6. The view with the list of items registered now has a new item. This item has the state of "Registered"
7. The user clicks in the item just created, and see the detailed view
8. In the detailed view we can see the QR of the good and some metadata. The item is in "Registered" state.
9. If we scroll down we can see the "Provenance" list with only one entry (green tick): "Registered"
10. If the user clicks in the first and unique item of the list a new view is opened showing the:
    • Digital signature of MSD as manufacturer
    • List of files associated with the good (product specs, quality report)
    • Each file will show a fingerprint (like the md5sum but abbreviated)
11. The user clicks in the "hand over" button
12. This open a view with the good QR code that can be shown to the DHL pick-up guy

DHL demo

The flow of the demo using a DHL user is:

1. We open the application as a DHL user
2. We show the DHL home page with all the goods managed by DHL and the different states (In transit, Delivered, Problem detected)
3. We open the "Pick up" button from the app, this open the camara view for scanning a QR
4. We scan the QR from the MSD app. We show a confirmation page where we say everything is looking good
5. We show again the home page, with the new cargo added, now the state is "In Transit"
6. The user clicks in the item just created, and see the detailed view
7. In the detailed view we can see the QR of the good and some metadata. The item is in "In Transit" state.
8. If we scroll down we can see the "Provenance" list with 2 entries (green tick): "Registered" and "Handed over"
9. If the user clicks in the second item of the list (Handed Over) a new view is opened showing the:
   • Digital signature of MSD as manufacturer and DHL as carrier
   • List of files associated with the good (product specs, quality report)
   • Each file will show a fingerprint (like the md5sum but abbreviated)
10. The user clicks in the "hand over" button

KLM demo

The flow of the demo using a KLM user is:

1. We open the application as a KLM user
2. We show the KLM home page with all the goods managed by KLM and the different states (In transit, Delivered, Problem detected)
3. We open the "Pick up" button from the app, this open the camara view for scanning a QR
4. We scan the QR from the DHL app. We show a confirmation page where we say everything is looking good
5. We show again the home page, with the new cargo added, now the state is "In Transit"
6. The user clicks in the item just created, and see the detailed view
7. In the detailed view we can see the QR of the good and some metadata. The item is in "In Transit" state.
8. If we scroll down we can see the "Provenance" list with 3 entries (green tick): "Registered" and "Handed over" two times
9. If the user clicks in the third item of the list (Handed Over) a new view is opened showing the:
   • Digital signature of DHL and KLM as carriers
   • List of files associated with the good (product specs, quality report)
   • Each file will show a fingerprint (like the md5sum but abbreviated)
10. The user clicks in the "hand over" button

Final recipient

The flow of the demo using the John Doe user is:

1. We open the application as a John Doe user
2. We show the user home page with all the goods controlled by the user and the different states (In transit, Delivered, Problem detected)
3. We open the "Pick up" button from the app, this open the camara view for scanning a QR
4. We scan the QR from the KLM app. We show a confirmation page where we say everything is looking good
5. We show again the home page, with the new cargo added, now the state is "In Transit"
6. The user clicks in the item just created, and see the detailed view
7. In the detailed view we can see the QR of the good and some metadata. The item is in "In Transit" state.
8. If we scroll down we can see the "Provenance" list with 4 entries (green tick): "Registered" and "Handed over" two times and "Delivered"
9. If the user clicks in the third item of the list (Handed Over) a new view is opened showing the:
   • Digital signature of KLM as carrier and John Doe
   • List of files associated with the good (product specs, quality report)
   • Each file will show a fingerprint (like the md5sum but abbreviated)
10. The user scroll down and can see the map with the complete journey of the cargo

Transit Temperature Acceptance (Add to KLM Demo)

The flow of the demo for this user story is:

1. KLM receives the House Way Bill from the Freight Forwarder (i.e. DHL) and scans the shipment's QR code
2. A digital copy of the House Way Bill is received upon scanning [assumption]
   • The Codes for Pharmaceutical, COVID (COV), Temperature Control (assumed) and HS Code for Vaccine and are flagged
3. The recipient is prompted to "Analyze Shipment Temperature?"
   • Selections to prompt are "Yes" and "No"
   • If "No" is selected, KLM takes receipt and signs transfer handshake
   • If "Yes" is selected, the following takes place:
4. KLM makes a request to Merck for access to their temperature unit data to determine if the temperature rose above X (KLM sets value) during any period prior to handoff
   • KLM access request is made to Merck
   • Payment for data access is made
   • KLM query (shipment ID, temperature threshold value) is sent to run on Merck data set
   • Query is run by Merck against temperature dataset
5. Based on query result, "Pass" or "Failed" response is returned to KLM (i.e. no source data is returned)
6. A prompt is displayed asking "Accept Shipment?"
   • Two option buttons are "Yes" and "No"
   • If "Yes" is selected, KLM takes receipt and signs transfer handshake
   • If "No" is selected, the package is returned to the freight forwarder

Customs Agent (Nice to Have)

The flow of the demo using the Customs Agent user is:

1. We open the application as a Customs Agent user
2. We show the user home page with all the goods needing clearance by the user and the in transit state
3. The Customs Agent wants to know who the Consigner is for a package so he presses the "Request Consigner Info" button
4. The Consigner, in this case MSD, gets a push notification that Customs is requesting additional information
5. The Consigner can select the "Accept Request" button, or the "Deny Request" button
   • If the Consigner selects "Accept", their infomation is signed and pushed the Customs Agent
   • If the Consigner selects "Deny", a "Request Denied" message is pushed to the Customs Agent
6. A response is received by the Customs Agent from the Consigner
7. The Customs Agent wants to know who the Consignee is for a package so he presses the "Request Consignee Info" button
8. The Consignee, in this case John Doe, gets a push notification that Customs is requesting additional information
9. The Consignee can select the "Accept Request" button, or the "Deny Request" button
   • If the Consignee selects "Accept", their infomation is signed and pushed the Customs Agent
   • If the Consignee selects "Deny", a "Request Denied" message is pushed to the Customs Agent
10. A response is received by the Customs Agent from the Consigner
11. The Customs Agent can then scroll down to two buttons: "Pre-Clear Shipment" or "Deny Pre-Clearance"
12. The Customs Agent selects either "Pre-Clear Shipment" or "Deny Pre-Clearance"
    • If "Pre-Clear Shipment" is selected, both the Consigner and the Consignee get a push notification that shipment has been pre-cleared by customs
    • If "Deny Pre-Clearance" is selected, both the Consigner and the Consignee get a push notification that shipment failed pre-clearance and additional information is required
13. The Customs Agent gets a notification that KLM is ready with a shipment and information about the shipment
14. The Customs Agent scrolls down to two buttons, one that says "Shipment Cleared" and what that says "Request Additional Information"
    • If the shipment matches a shipment that has been pre-cleared, the user selects the "Shipment Cleared" button
    • If the shipment does not match a shipment that has been pre-cleared, the user can request the Consigner and Consignee information (see same flow above)
      • The user selects either "Approve Shipment" or "Deny Shipment"
      • If "Approve Shipment" is selected, the Consignee, the Consigner and KLM get push notifications that shipment has been "Approved for Shipping" by customs
      • If "Deny Shipment" is selected, the Consignee, the Consigner and KLM get a push notification that shipment failed clearance and additional information is required

Bad Handover scenarios

Since the handover seems to be the most important part of the supply chain it is important to go through some of the failure scenarios that can happen during the handover and detail how our solution behaves in this case. The users of this solution don't want just to see an ❌ but we want to also know what was the reason for the handover to fail.

Imagine the case where a handover fails between MSD and DHL. Which side is at fault? Maybe DHL presented wrong identity credentials during the handover and MSD decied to cancel the handover. Or maybe DHL after looking at the metadata associated with the cargo found problems and decided to cancel the handover.

In this section we detail some failed handover scenario. We will focus on two actors, with MSD trying to handover cargo the DHL.

Scenario 1 - Failed due diligence

1. MSD clicks in the "hand over" button:
   • MSD possibly needs to select the identity of DHL or this is already somehow predefined for this cargo.
   • This represents MSD's intent to handover the cargo to DHL
   • (Could this action be recorded in the provenance contract?)
2. This open a view with the QR code (unique to this handover) that can be shown to the DHL pick-up guy
3. DHL opens the application
4. We show the DHL home page with all the goods managed by DHL and the different states (In transit, Delivered, Problem detected)
5. DHL clicks the "Pick up" button from the app, this open the camara view for scanning a QR
6. We scan the QR from the MSD app. We show a confirmation page where we say that there is a problem with the cargo
7. We show again the home page, with the new cargo added, still in the state "Registered"
8. The user clicks in the item just created, and see the detailed view
9. In the detailed view we can see the QR of the good and some metadata. The item is in the "Registered" state.
10. If we scroll down we can see the "Provenance" list with 2 entries (green tick): "Registered" and (greyed out pending) "Handed over"
11. If the user clicks in the second item of the list (Handed Over) a new view is opened showing the:
    • Digital signature of MSD as manufacturer
    • List of files associated with the good (product specs, quality report)
    • Each file will show a fingerprint (like the md5sum but abbreviated)
12. DHL clicks in the "cancel hand over" button
    • Maybe this can also be recorded in the provenance contract

Depending on what is recorded in the provenance contract in this scenario we could learn that:

• MSD met with DHL and showed intent to handover the cargo
• DHL denied the handover specifying some reason

Dev quickstart

• Clone git
• Install expo framework globally using npm install -g expo-cli
• Install packages (inside folder) yarn
• Start web livereload app with expo start --web

Nevermined integration

The dashboard is available here: https://dashboard.keyko.rocks/

We are using Rinkeby Ethereum network for keeping track of the provenance record:

Services

• Keeper (Infura) - https://rinkeby.infura.io/v3/1c75f192257c497e91f30832666d179e
• Gateway - https://gateway.keyko.rocks/
• Metadata API - https://metadata.keyko.rocks/

Accounts

Owner	Address	Path
MSD	0xe2DD09d719Da89e5a3D0F2549c7E24566e947260	0
DHL	0xBE5449a6A97aD46c8558A3356267Ee5D2731ab5e	1
KLM	0xA78deb2Fa79463945C247991075E2a0e98Ba7A09	2
John Doe	0x02354A1F160A3fd7ac8b02ee91F04104440B28E7	3
Customs	0xe17D2A07EFD5b112F4d675ea2d122ddb145d117B	4

The mnemonic is: taxi music thumb unique chat sand crew more leg another off lamp

Assets

For more information about the assets to register and use during the demo and the QR codes, please go to the Assets documentation page.

Links

• Jerry Figma file - Initial prototype
• Esraa Figma file - Prototype implementing the flow described here
• Odyssey Hackathon presentation

odyssey-dashboard's Issues

Style Dashboard/homepage

Implement the KLM user flow

Integrate the QR & camara in the application

Implement the DHL user flow

Review the files to use during the demo

Deploy the app in K8s

Implement the John Doe user flow

Collect all the accounts

Provide the library to connect with Nevermined

Integrate the app with the Nevermined Provenance record

Customs Wireframes

Apply the wireframe styles to the app

Deploy Nevermined on Rinkeby

Implement the customs office user flow

Implement the MSD user flow