Project Ordö takes inspiration from history, biology, and artificial intelligence to help train autonomous vehicles (AVs) to be safe and welcoming for the members of each community they enter.


First Principles

Mutual Prosperity

All participating stakeholders may enjoy the value of the data generated by this project via open-source and data-trust models.

Excellence in Safety

We believe that anybody’s accident is everybody’s accident. Like NASA and the National Transportation Safety Board (NTSB), our approach to the prevention of catastrophic failures is to develop technological and human systems that surface matters of fact after an error occurs rather than to assign blame.


Our systems are designed to empower those whose quality of life will be disrupted first and/or worst by the emergence of autonomous transportation.

An All-Party Initiative to Maximize Safety Throughout "The First 500 Accidents"

No one can predict with certainty the full set of potential errors that AVs will encounter when put to use on public roads. To avoid preventable harms, foster public trust, and enable innovation, we propose to have all-parties agree to share data related to the first 500 incidents involving an AV and a fatality. As part of this pact, stakeholders would benefit from blameless, fact-driven accident reporting as well as certain liability and intellectual property protections. This proposal builds on the wildly successful model used by the NTSB in aviation, under which "Anybody’s accident is everybody’s accident" and on the International Space Station, in which "One mistake is everybody's mistake".

"Community Mode": 21st Century Accident Reporting to Share Responsibility Between Industry and the Public

The creation of a simple platform via which the public can express and democratically consolidate their visions for how AVs can improve their community.

A Safety Certification Program: A Driving School for AVs

What should a driving test for AVs look like? How do we offer multiple stages of certification considering the complexity of the technology and the difficulty of the driving task in real world environments. We propose a peer review process that aims to provide asnwers to these questions by utilizing the growing field of Causality and Bayesian Inference. Our safety certification program builds on the idea that innumerable accidents can be prevented over the long term by having different AV products compete against each other to pass certain tests.

AVs Reliability Monitoring

Shared Prosperity: An "Inclusion Rider" to Hire Those Whose Jobs Will be Most Impacted by the Adoption of AVs

Evidence suggests that marginalized groups like non-English speakers in the U.S. will be hit hardest by the adoption of autonomous vehicles, since they are often employed to a disproportionate degree in the taxi and trucking industries, which each face considerable risk of disruption. As good community members, the AV industry could employ an "inclusion rider," which is roughly (in this context) a promise to hire those who have been impacted most by the incoming technology. Using the system outlined above, marginalized groups could be re-employed to participate in the development of the safety-infrastructure we have just proposed. Little training will be needed, for instance, to help propagate the Shared Semantic Space that enables "Community Mode" with useful, common-sense tags and categories that help to predict and prevent accidents. In NYC, no one would know this data better than cab drivers. The resulting data-set could be organized as a data-trust, which would ensure that contributors participated in revenue gained for its exploitation via a co-operative license.


Jonnie Penn
Jonnie Penn, from History and Philosophy of Science Department at the University of Cambridge, studies the impact that artificial intelligence and other emerging technologies have on society.
Bogdana Rakova
Engineer and ML researcher at Think Tank Team, Samsung Research America. Previously had a startup in the manufacturing field. Working on tools that empower people to make better decisions. AI Safety Geek.

Let's Get In Touch!

Interested in the space? Give us a call or send us an email and we will get back to you as soon as possible!