←back to thread

Launch HN: Flywheel (YC S25) – Waymo for Excavators

103 points jashmota | 2 comments | 24 Sep 25 16:48 UTC | HN request time: 0.017s | source

Hey HN, We're Jash and Mahimana, cofounders of Flywheel AI (https://useflywheel.ai). We’re building a remote teleop and autonomous stack for excavators.

Here's a video: https://www.youtube.com/watch?v=zCNmNm3lQGk.

Interfacing with existing excavators for enabling remote teleop (or autonomy) is hard. Unlike cars which use drive-by-wire technology, most of the millions of excavators are fully hydraulic machines. The joysticks are connected to a pilot hydraulic circuit, which proportionally moves the cylinders in the main hydraulic circuit which ultimately moves the excavator joints. This means excavators mostly do not have an electronic component to control the joints. We solve this by mechanically actuating the joysticks and pedals inside the excavators.

We do this with retrofits which work on any excavator model/make, enabling us to augment existing machines. By enabling remote teleoperation, we are able to increase site safety, productivity and also cost efficiency.

Teleoperation by the operators enables us to prepare training data for autonomy. In robotics, training data comprises observation and action. While images and videos are abundant on the internet, egocentric (PoV) observation and action data is extremely scarce, and it is this scarcity that is holding back scaling robot learning policies.

Flywheel solves this by preparing the training data coming from our remote teleop-enabled excavators which we have already deployed. And we do this with very minimal hardware setup and resources.

During our time in YC, we did 25-30 iterations of sensor stack and placement permutations/combinations, and model hyperparams variations. We called this “evolution of the physical form of our retrofit”. Eventually, we landed on our current evolution and have successfully been able to train some levels of autonomy with only a few hours of training data.

The big takeaway was how much more important data is than optimizing hyperparams of the model. So today, we’re open sourcing 100hrs of excavator dataset that we collected using Flywheel systems on real construction sites. This is in partnership with Frodobots.ai.

Dataset: https://huggingface.co/datasets/FlywheelAI/excavator-dataset

Machine/retrofit details:

  Volvo EC380 (38 ton excavator)
  4xcamera (25fps)
  25 hz expert operator’s action data
The dataset contains observation data from 4 cameras and operator's expert action data which can be used to train imitation learning models to run an excavator autonomously for the workflows in those demonstrations, like digging and dumping. We were able to train a small autonomy model for bucket pick and place on Kubota U17 from just 6-7 hours of data collected during YC.

We’re just getting started. We have good amounts of variations in daylight, weather, tasks, and would be adding more hours of data and also converting to lerobot format soon. We’re doing this so people like you and me can try out training models on real world data which is very, very hard to get.

So please checkout the dataset here and feel free to download and use however you like. We would love for people to do things with it! I’ll be around in the thread and look forward to comments and feedback from the community!

Show context
seabrookmx ◴[24 Sep 25 19:04 UTC] No.45364639[source]▶
> The joysticks are connected to a pilot hydraulic circuit, which proportionally moves the cylinders in the main hydraulic circuit which ultimately moves the excavator joints

I've actually spent a decent amount of time running an excavator, as my Dad owns a construction / road building company. It was a great summer job!

An important note about the pilot hydraulics is that they _provide feedback to the operator_. I would encourage any system that moves these controls on behalf of a remote human operator or AI to add strain gauges or some other way to measure this force feedback so that this data isn't lost.

The handful of "drive by wire" pieces of equipment that my Dad or other skilled operators in my family have ran were universally panned, because the operators are isolated from this feedback and have a harder time telling when the machine is struggling or when their inputs are not sufficiently smooth. In the automotive world, skilled drivers have similar complaints about fully electronic steering or braking systems, as opposed to traditional vacuum or hydraulic boosting approaches where your foot still has a direct hydraulic connection to the brake pads.

replies(4): >>45365403 #>>45365849 #>>45366303 #>>45366966 #
jeffbee ◴[24 Sep 25 20:11 UTC] No.45365403[source]▶
My car with its drive-by-wire brakes has a brake feedback simulator that gives the driver the kind of feeling associated with power-boosted hydraulic brakes. This is by far the most expensive single component in the car. Arguably these are just expensive accommodations for human flaws. A self-driving car wouldn't need them. Can't the self-driving system act directly on data like pressure, flow, and displacement?
replies(3): >>45365470 #>>45366607 #>>45366959 #
1. 01HNNWZ0MV43FF ◴[24 Sep 25 20:18 UTC] No.45365470[source]▶
Maybe it doesn't matter for a car because feeling the car's motion tells you most of what you need to know. A car is not meant to touch anything but the road, in normal conditions. I think steering is the only case where force feedback is very important for a car - In the winters up here, I can feel the steering go loose when I hit a patch of ice.

I imagine an excavator, meant to touch and dig through things, and lift things, benefits from force feedback for the same reason VR would.

Have you played those VR sword games? BeatSaber works great because you're cutting through abstract blobs that offer no resistance. But the medieval sword-slashing games feel weird because your sword can't impact your opponent.

I saw a video recently of a quadcopter lifting heavy objects. When it's overloaded, it can't maneuver because all its spare power is spent generating lift to maintain altitude. If the controls had force feedback, the copter's computer could tell you "I'm overloaded, I can't move" by putting maximum resistance on the sticks.

replies(1): >>45366682 #
2. jashmota ◴[24 Sep 25 22:25 UTC] No.45366682[source]▶
Interestingly, we had some people try out VR teleop: https://x.com/Scobleizer/status/1970245161306464667

https://x.com/jash_mota/status/1969091992140304703

I think force feedback is key for small excavators, but not really true for 25+ tons excavators. Hence how easy it is for operators to accidentally kill someone with it.