Imagine a squad of autonomous drones searching for smoke as they fly high above the Sierra Nevada mountains, armed with superior sensor equipment. When these leader robots detect a wildfire, they send instructions to a swarm of firefighting drones, which race to the blaze’s location.
But what if a bad actor hacked one or more leader robots and started delivering erroneous directions? How would the follower robots know they’d been deceived as they were brought further away from the fire?
According to a study published today in IEEE Transactions on Robotics by researchers at MIT and the Polytechnic University of Madrid, using blockchain technology as a communication medium for a team of robots could provide security and safeguard against deceit.
The discovery could be useful in places where multi-robot self-driving car systems are delivering commodities and transporting people around.
A blockchain provides a tamper-proof record of all transactions, in this case, the messages sent by robot team leaders, allowing follower robots to spot discrepancies in the data trail.
According to Eduardo Castelló, a Marie Curie Fellow in the MIT Media Lab and lead author of the paper, leaders use tokens to signal movements and add transactions to the chain, and they forfeit their tokens when they are caught in a lie, so this transaction-based communications system limits the number of lies a hacked robot could spread.
“The world of blockchain beyond the discourse about cryptocurrency has many things under the hood that can create new ways of understanding security protocols,” Castelló says.
Since we know how lies can impact the system and the maximum harm that a malicious robot can cause in the system, we can calculate the maximum bound of how misled the swarm could be. So, we could say, if you have robots with a certain amount of battery life, it doesn’t really matter who hacks the system, the robots will have enough battery to reach their goal.Eduardo Castelló
Not just for Bitcoin
While a blockchain is most commonly associated with cryptocurrency, it is essentially a series of data structures known as blocks that are linked together in a chain. Each block contains the data it is intended to keep, as well as the “hash” of that data and the “hash” of the previous block in the chain. The process of transforming a string of text into a set of unique numbers and letters is known as hashing.
The information stored in each block in this simulation-based study is a series of directions from a leader robot to followers. If a malevolent robot tries to change the content of a block, the block hash is changed, and the changed block is no longer attached to the chain.
Follower robots may be oblivious to the changed routes. In addition, the blockchain keeps a permanent record of all transactions. Because all followers will eventually be able to observe all of the directives given by leader robots, they will be able to determine if they have been deceived.
For example, if five leaders transmit messages instructing followers to walk north but one leader instructs followers to move west, the followers may disregard the inconsistency.
Even if a follower robot moved west by accident, the deceived robot would soon recognize the error when comparing its movements to the blockchain transactions.
Each leader in the researchers’ system is given a set number of tokens, which are used to add transactions to the chain. Only one token is required to make a transaction. The leader loses the token if followers find that the information in a block is untrue by examining what the majority of leader robots signaled at that particular stage. When a robot’s tokens run out, it can no longer deliver messages.
“We envisioned a system in which lying costs money. When the malicious robots run out of tokens, they can no longer spread lies. So, you can limit or constrain the lies that the system can expose the robots to,” Castelló says.
The researchers put their approach to the test by simulating numerous follow-the-leader scenarios with a known or an unknown number of malevolent robots. Leaders broadcast wrong directions or sought to block the path of follower robots using a blockchain, while bad leaders broadcast false directions or attempted to block the path of follower robots.
Even when follower robots were first misled by hostile leaders, the transaction-based approach enabled all followers to eventually reach their goal, according to the researchers. The researchers created techniques to establish the maximum amount of falsehoods a hostile robot can tell because each leader has an equal, finite number of tokens.
“Since we know how lies can impact the system, and the maximum harm that a malicious robot can cause in the system, we can calculate the maximum bound of how misled the swarm could be. So, we could say, if you have robots with a certain amount of battery life, it doesn’t really matter who hacks the system, the robots will have enough battery to reach their goal,” Castelló says.
The algorithms enable a system designer to estimate the amount of battery life required for the robots to complete their task, as well as the amount of memory required to store the blockchain, the number of robots required, and the length of the path they can travel, even if a percentage of the leader robots are hacked and become malicious.
“You can design your system with these tradeoffs in mind and make more informed decisions about what you want to do with the system you are going to deploy,” he says.
Castelló plans to build on this work in the future to develop new security methods for robots based on transaction-based interactions. He sees it as a means of fostering trust between humans and robot groupings.
“When you turn these robot systems into public robot infrastructure, you expose them to malicious actors and failures. These techniques are useful to be able to validate, audit, and understand that the system is not going to go rogue. Even if certain members of the system are hacked, it is not going to make the infrastructure collapse,” he says.
The paper was co-authored by Ernesto Jiménez and José Luis López-Presa of the Universidad Politécnica de Madrid. The European Union’s Horizon 2020 Research and Innovation Program, the Madrid Regional Government, and the MIT International Science and Technology Initiatives Global Seed Fund all contributed to this study.