stefan8r

9 ms·

1.
Polymath Robotics (YC S22) Is Hiring(ycombinator.com)
1 points by stefan8r 3y ago
2.
by stefan8r 4y ago
I seem to recall a time when they all left earth and thanked us for all the fish...
3.
by stefan8r 4y ago
In fact, during our interview with TechCrunch we let Kirsten drive our unmanned test vehicle from her browser at home!
4.
by stefan8r 4y ago
I'd challenge this assertion. Rio Tinto has been on the forefront of automation, but there's a long list of equipment they operate that they don't have an immediate pathway to automate. To my knowledge, the main players auto
5.
by stefan8r 4y ago
I always liked this reference as I thought it applied to a lot of ML stuff. Step 1: Collect Data, do some analysis, train a model or two Step 2: … Step 3: Profit!
6.
by stefan8r 4y ago
Thanks so much! Glad to see you’re doing well!
7.
by stefan8r 4y ago
Spot on. Except we’re just the onboard software (BYO HW)
8.
by stefan8r 4y ago
Thanks - appreciate that. And thanks for reading my response in the right tone (I was nervous no inflection might do me a disservice). I’m not familiar with Tangram, but thanks for sharing. I’ll take a look! There is definitely a world wh
9.
by stefan8r 4y ago
Thanks so much for the kind words! 1) We are mostly built on ROS (for better and worse) and are starting to migrate some containers to ROS2. Down the road I see an architecture that has some ROS, some ROS2, and maybe some homemade stuff (o
10.
by stefan8r 4y ago
To double click on the Twilio API example: In most robotics applications teams are building the equivalent of new speculative phone networks, building specialized infrastructure, acquiring specialized hardware companies to ease computer:pho
11.
by stefan8r 4y ago
Might be a bit of a misunderstanding here. We’re not providing a sim world for you to do ML training on. Caladan would actually be a pretty terrible tool for that - it’s a static low-res environment without any other agents. In Caladan we
12.
by stefan8r 4y ago
Thanks for sharing Jaybridge. I hadn’t come across them but will reach out / try to dig into them.
13.
by stefan8r 4y ago
Thanks so much for the note and the kind words. Would love to grab a coffee or something! My email is first @ polymathrobotics.com, but I'll reach out as well! Maybe we can play bocce instead of coffee?
14.
by stefan8r 4y ago
This is a good, well thought out point that I 60% disagree with (respectfully, of course). I think you're absolutely accurate in describing how robotics is today, and you're definitely partially correct about how it will remain,
15.
by stefan8r 4y ago
Ya video summary is nothing to sneeze at - I'll take my ML tasks as easy as determining if something is a person or a pixel /s Completely agree on decomposing hard things.
16.
by stefan8r 4y ago
Ish. To _probably_ annoy you and everyone on this thread, I'll tell you how this works for a "large publicly traded company that has autonomous vehicles in mines." Essentially they have a portion of the mine site designated
17.
by stefan8r 4y ago
To add to what Ilia said - re: - "The machine works good enough that it can perform the task until it can't anymore and then someone remotes-in to fix it?" The machine will work good enough for basic autonomy, but then the re
18.
by stefan8r 4y ago
Ilia gave you the more real (and specific answer) but I just want to call out that I love the Underpants Gnome reference. I used to cite it a lot and was frustrated by how many times I had to explain it. Sim =/= sim =/= sim. By
19.
by stefan8r 4y ago
To add a somewhat easy sentance here - we're our actual product is delivered as SaaS (and it's not at Gmail pricing, it's more enterprise). Were you to become a customer we'd be pretty handholding with you to get the th
20.
by stefan8r 4y ago
Thanks so much! We had a fantastic designer, and our marketing/ops/cust success/everything else lead is really way better than we deserve!
21.
by stefan8r 4y ago
You should get off the sidelines! Post-Starsky I wanted to run away from HardTech / FrontierTech and just do "easy SaaS stuff." I was deeply upset to realize I was too broken to do something that "easy." (still, re
22.
by stefan8r 4y ago
Thanks for the kind words! Ya - I can't imagine how hectic it was for you guys. I remember hearing Josh Switkes talk about hand building radars for pre-2010 autonomy. Re:booms/buckets/etc: right now we're just focusing
23.
by stefan8r 4y ago
Thanks! It would be lovely if posting this one time in HN single handedly kicked off that ecosystem -> so everyone reading should quit their jobs and build robots /s More seriously - there are a surprisingly large number of groups
24.
by stefan8r 4y ago
To add to this - a neat thing about swarm robots is splitting the compute load between robots and figuring out how different robots move in conjunction with eachchoter. We're just focusing on the software that sits on the robot itself
25.
by stefan8r 4y ago
And, I should say, thanks for the kind words!
26.
by stefan8r 4y ago
We are using Gazebo for working on controls (getting Caladan ready to launch also led to a number of big improvements in controls) but not for ML training. The point of Caladan is more for you to be able to build specific autonomy functiona
27.
Launch HN: Polymath Robotics (YC S22) – General autonomy for industrial vehicles
172 points by stefan8r 4y ago|73 comments
28.
Robotics is about to (finally) explode(medium.com)
3 points by stefan8r 4y ago|0 comments
29.
Throwing Wrenches at Autonomous Vehicles: an 80-20 rule test for viability(forbes.com)
1 points by stefan8r 6y ago|0 comments
30.
by stefan8r 6y ago
They can't. Trucking is an undifferentiated / commoditized service in a fragmented market. If one company increases pay to drivers, they can't increase rates to customers because they'll just shift their demand to some

More ›