Causality Network: IoT driven data collection in science

Title: Causality Network: IoT driven data collection in science

Team member names: Bharat and Erin

Short summary of your improvement idea
Causality Network seeks to revolutionize the data collection protocol in scientific experiments. This project aims to address the reproducibility crisis in science due to research vulnerabilities from data tampering and falsification by adding a blockchain-based layer onto IoT device data generation in research experiments. This protocol addition will be field tested with open-source hardware in research investigations.

Q&A
What is the existing target protocol you are hoping to improve or enhance?
How data is collected in scientific experiments from IoT devices.

What is the core idea or insight about potential improvement you want to pursue?
IoT devices are at the core of scientific research from wet lab research (e.g. bioreactors) to clinical trials (e.g. blood test machines). These processes are increasingly automated via robotics.

There are weaknesses in the current system with the ability to easily tamper data or falsify it (the latter is a particular concern with AI being able to falsify large data sets that appear authentic). The inability to provide raw data is a major factor contributing to the reproducibility crisis in science.

We are working on a protocol that leverages blockchain and secure cryptographic methods (like zK proofs and secure enclave signatures) to not only have an audit trail of data but proof of its authenticity.

What is your discovery methodology for investigating the current state of the target protocol?
For transparency/context:
Bharat is the co-founder of DeSci London, which has run monthly meetups since 2022 and organises annual conferences. It has the largest IRL DeSci community (over 700 academics).

Erin and Bharat are co-founders of MuseMatrix, a DeSci Fellowship taking 20 fellows (10 web3 developers and 10 scientists) through a part-time six month programme. The first two months are spent on a curriculum and then fellows work collaboratively to solve one big problem.

This cohort will focus solely on Causality Network. The scientists will be from a variety of fields where IoT devices are common (eg SynBio, NeuroTech, BioSecurity) and will therefore have familiarity with the current state of IoT driven research.

We have received a small amount of funding from the Ethereum Small Grants Programme for technical development. We already have a development framework/design for the protocol.

Money from this grant will be used as a stipend allowing us to work full time on this project. This will also mean we can conduct interviews and speak to more experts/researchers in the field.

In what form will you prototype your improvement idea?

A prototype application that can integrate different IoT devices to run scientific experiments. Aimed at scientists that are not familiar with blockchain.

How will you field-test your improvement idea?
MuseMatrix Fellows will develop a proof of concept with the protocol running on open-source IoT devices, conducting scientific experiments and generating reports for community feedback.

An example of a device/experiment is The Pioreactor (a bioreactor built on top of a Raspberry Pi). It’s easy to modify and has a preexisting developer community. We can run basic experiments (e.g. brewing alcohol, glow in the dark bacteria) and test our improvement idea. There are similar plans with other IoT equipment the Fellows are familiar with (such as Open BCI).

We will then implement this process in a partner lab (e.g. ChimiaDAO’s chemistry lab in Colombia) to ensure end-to-end integration.

Who will be able to judge the quality of your output?

  • Scientists for usability in their research (esp those not familiar with blockchain)
  • Technical mentors working with the fellowship for protocol analysis and improvements
  • Partner labs for practical feedback on integration.
    (We will add specific names in the upcoming days)

How will you publish and evangelize your improvement idea?.

  • We will use a community centric approach.
  • We will utilise DeSci London monthly meetups/ conferences to hold talks and workshops, raise awareness, and garner feedback.
  • We will also hold interactive hands-on workshops in Chicago and ZuGeorgia (Zuzalu inspired event), & submit presentations at other global DeSci events (eg Devcon in Bangkok)
  • We are partnered with a number of DeSci DAOs (such as ValleyDAO) and other DeSci communities (such as DeSci Asia).
  • Results will be published on DeScier (a web3 DeSci journal) and Paragraph
  • Code will be published open source

What is the success vision for your idea?

The ultimate success for Causality.Network is becoming the standard protocol in scientific research for creating trustworthy data. This would require 1. recognition of the increased data integrity & security provided in this protocol and 2. adoption with multiple IoT devices in science.

5 Likes

Very interesting to see this. We’re working on a different but related problem, and share the conviction that people creating their own custom instruments are the natural ground for new approaches to data traceability and forensics.

It sounds like you are already embedded in academic research networks, so have experience with the insane depth of customization involved in sensor-based research. Almost every step of the process—from development of hardware, low-level software that controls things like how much energy is used and how data is stored and transferred, to routines of observation such sampling rate and calibration, to data collection routines (often involving human actors, not all of them perfectly functional), to aggregation methods, etc. etc. etc. — has the potential for customization, and some parts of this system are virtually certain to be customized in any research lab. The friction involved in inserting new technologies into the research stack is correspondingly high. People who watch from the outside often mistake standardization of the output in the form of research papers as representing some kind of standardization of the inputs in the observational process, but this is an illusion. If you have time for amusement, the history of C.C. Little and Jackson Labs, in Maine, where they produce standard mice, is an interesting excursion into some of the paradoxes of imposing standard form on scientific inputs.

Will be very glad to follow this project as it develops.

Thank you for the reply and your kind words! :grinning:

I’m the team member for this project, nice to e-meet you!

Sounds like you have quite a bit of academic experience @agaricus?

You’re completely right, the complexity in senor labs is an absolute nightmare, and trying to add something new to the stack can be virtually impossible. We want to support people using whatever tool they can and in a frictionless way allow that data to be validated from the instance it is generated.

There is some hope in getting things a bit more standardised as there is a trend with more labs (in London/Oxford/Cambridge at least) using open-source automated equipment such as Opentrons meaning you can customise a standardised bit of hardware

I’m super interested in rentable open lab spaces (separate point but I believe the creator economy for scientists is around the corner where scientists will be able just to rent a lab
and own the IP of what they work on). Open.Cell did this but they have recently closed down. Even more wild is Emerald Cloub Labs which is completely automated and remote so places like this would be a natural fit for us to begin with.

There was a recent conference on AI/Automated labs if you are interested, lots of good content.

My dream is to buy a shipping container and retrofit it with completely AI driven automated equipment :slight_smile:

Will check out “Making Mice” - thanks :slight_smile:

2 Likes