Sharing weather data autonomously using Fetch.ai – Part 3
Nov 25, 2019
Fetch.ai is using the example of data being shared between a weather station and an individual to show how knowledge is traded on the network.
In the first article in this series, we discussed how agents find, connect and interact with each other. In our second post, we examined how agents transact across the Fetch.ai smart ledger. In this article, we will take a look at some of the trust issues agents face when completing transactions using the ledger, and some of the solutions we have implemented to address these challenges.
As stated in our second article, the balance of trust between Charlotte’s agent (which is seeking weather data on her behalf) is not equal to that of the agent representing the weather station. This is because the weather station agent only shares its data after it has received the micropayment from Charlotte. For this reason Charlotte’s agent needs to trust that the weather station will act honourably.
The traditional way to resolve this issue is to create an escrow transaction with a third party. This would mean that the third party would securely hold the micropayment and the data until both had been verified. Once this process was complete, the transaction would take place. In reality though, such an approach is impractical as the third party would have to be properly contracted and its involvement would inevitably increase the overall transaction cost of the trade. To overcome this problem, the agents could use an escrow smart contract. However, this solution doesn’t address the possibility that the weather station may provide inaccurate or irrelevant data, both of which would leave Charlotte shortchanged.
For the trading of weather data, the most obvious solution is for the weather station to have a trust rating on the ledger. A trust rating enables other agents to see whether an agent has a good track record of honouring its obligations. A high rating suggests that, based on previous behaviour, there is a strong chance the agent will send the data it is advertising. A low trust rating indicates the agent has previously failed to fulfill its side of the bargain. This information would be accessible to all agents on the Fetch.ai network.
Each agent looking to trade would seek to conduct a transaction with an agent that has a high trust rating. Therefore, a poor reputation will damage an agent’s ability to trade on the Fetch.ai network. So too does a blank record. Just as you may be understandably more wary about staying in accommodation on Airbnb which hasn’t had any previous visitors, compared with a place that has lots of positive reviews, agents are also less likely to conduct economic activity with agents that do not have an extensive track record of acting honestly. This detail is important as it prevents malevolent agents from simply creating a new identity after each dishonest transaction. Agents that do have a bad trust rating are, over time, recognised as such and are expelled from the Fetch.ai network.
- If you’re a developer, find out more about Fetch.ai’s Autonomous Economic Agents by checking out the documentation section of our website, which includes some demos.
- Watch videos outlining our technology and follow our tutorials on YouTube.
- Join our Developer Slack Channel and explore our growing GitHub repository.
- If you’re not a developer, stay up to date with our progress by joining us on Telegram or follow us on Twitter.