Milestone: Droids Feasibility

The goal of Droids is ambitious. We want to take a nervous system and let anyone build smart mobile robots (droids) that their grandma can program.

The first question we asked ourselves, is whether our vision (🤖Droids for Everyone) was feasible. To that end we have:

read over 1,200 papers since June of 2023, with 80% of those in 2024.
canvased all the major companies from Tesla to Agility Robotics.
looked at every layer of the stack from hardware to software to ML
built or replicated dozens of hands on proof-of-concepts for core ideas

We regret this dev log did not exist sooner for folks to follow along in more detail.

However, our conclusion is that it is feasible and that, after waiting since 2019, now is the exact right moment to move and ride a rising tide of numerous technologies laid out in Commoditizing the Newton.

Here are some of the key questions we asked along with a summary of findings.

‣

Is there a need? Is this an Apple M1 moment? Is it time for “garage robotics”? Yes

‣

Is there significant moral upside? Yes

‣

Is there significant financial upside? Yes

‣

Is there an opportunity for at least 10x better tech? Yes

‣

Are the ML techniques good enough for sim2real? Yes

‣

Will the flywheel work? Can you use the same learning system across unlimited robot bodies? Yes

‣

Is there a killer interface anybody can instantly understand? can we make the killer interface available? Yes

‣

Is the hardware cheap enough to make this accessible? Yes

‣

Is DIY a big enough market to start? Yes