Case study

Zynoptes

How Zynoptes launched a simulation engineer for robotics to 190K+ views

Robots fail on site in ways nobody tested for, and every failure costs a truck roll. Zynoptes finds those failures in simulation first. The launch film had to make robotics engineers feel that problem before it showed them the fix.

Views on X
Likes on X
260+
Bookmarks on X
40

Zynoptes × EPTA

Zynoptes is an autonomous simulation engineer for robotics. Its CLI, ZynoCode, builds a simulation of your robot and its site, runs it through hundreds of variations, finds the failures and fixes them in your code, before the robot ships.

Owen Bales and Max Kern founded it, and it launched from Alliance's ALL18 cohort.

The brief

Simulation is the kind of product robotics engineers already believe in and rarely have time for. Building each world by hand, allocating compute, waiting on runs, fixing one bug and watching the robot break again: the cost is familiar, so a film that opened on features would read as one more tool to set up.

The film had to start on site, with the failure, and only then show how little the engineer has to do to prevent it.

The film

It opens on a solar panel cleaning robot sliding off the edge of its panel on a solar farm. On a real site that means someone drives out, fixes it and tests it again. Then the reveal: Zynoptes built that simulation, caught the robot failing, and fixed the code before it ever happened on site.

The first shot is the failure. The product only arrives once you have watched it cost someone a morning.

Owen and Max carry the rest on camera. Max starts with a task in plain English, Zynoptes builds the solar farm panel by panel, and the film runs the variations engineers never get to: dew, tilt, dust, low sun. Four hundred worlds on sixteen machines in forty-four minutes, three root causes traced, the fixes passing every new world they were tested on. One minute and thirty-four seconds, closing on the line from the post: break it before the world does.

Next launch

Planning a launch?

Tell us what is launching and when. Thirty minutes is usually enough to know whether we are the right team for it.