Robotics Software Engineer

• Develop and ship ROS2 features that enhance machine behavior and unlock new capabilities
• Build out our fleet management layer - telemetry, OTA, remote ops, fleet-wide observability
• Work on perception, sensor fusion, and decision-making in unstructured outdoor environments
• Collaborate directly with mechanical, electrical, and controls engineers - and with the technicians who validate and field-test what you build
• Make architectural calls that affect what the platform can do two and three years out
• Help shape how we hire, review, and scale the software team as we grow, * Outdoor or off-road robotics experience (agricultural, construction, defense, mining, AMR)
• Production computer vision in unstructured environments
• OTA update systems, fleet telemetry pipelines, remote operations tooling
• Embedded Linux, real-time systems, or low-level controls experience
• GPS-denied or GPS-degraded localization
• Work that’s been deployed at meaningful scale (hundreds-plus units)

ABOUT THE TEAM

The engineering team is ten people. No one is coasting.

We don’t have process layers between you and the machine. Code review matters here. Design discussions are technical and direct. Decisions get made by the people closest to the work - and the company is led by an engineer who knows the difference between cheap and right. We make trade-offs deliberately, not by default.

You’ll have unusual ownership for a company our size, and peers who’ll push back hard when you’re wrong.

We’re based near Green Bay, Wisconsin. Cost of living is a fraction of what you’d pay in a coastal tech hub, the commute is ten minutes, and the work is on real robots in real terrain - not a campus parking lot.

WHAT THIS IS NOT

A research position. A role where you’ll spend a year on a prototype that doesn’t ship. A place where someone else writes the spec and you implement it. A team where senior means “has been here longest.”

• Strong production ROS2 experience - not a class, not a side project; you’ve shipped and maintained ROS2 software that users depend on
• Solid C++ and/or Python; comfort moving between them
• Experience with at least one of: fleet management/distributed systems at scale, or computer vision and sensor fusion in real-world conditions
• A sense of when to build, when to integrate, and when to throw something out
• The judgment to operate as a senior IC: you scope your own work, push back when the spec is wrong, and ship without needing to be managed
• Clear technical writing - design docs, postmortems, code review comments that move the team forward

We want to know:

• What was the problem?
• What did you build?
• What broke, and how did you find it?
• What would you do differently?
• What did you learn that changed how you work?

Code, design docs, postmortems, talks, papers, demos - anything that shows how you think.

ABOUT RC MOWERS

We design and build autonomous and remote-operated mowing equipment for difficult environments - a growing, profitable company shipping to municipalities, DOTs, and commercial contractors across North America.

COMPENSATION & BENEFITS

Compensation commensurate with experience. Benefits include health, dental, and vision insurance, 401(k) with company match, paid vacation and sick leave, and disability, life, and accident insurance.

Hiring is subject to a pre-employment background check and drug screening. RC Mowers provides equal employment opportunities to all employees and applicants.

Within that, two areas have outsized leverage on where the platform goes - and you’d help us decide where to push hardest:

The fleet problem. We’re scaling toward thousands of units operating in the field. That means treating robots more like servers - telemetry, observability, OTA updates, fleet-wide health, remote diagnostics, mission orchestration, the whole stack that lets a small team operate a large fleet without things falling apart. If you’ve done backend or distributed systems work and want to bring that discipline to robotics, this is where it shows up.

The perception problem. LiDAR is great until it isn’t. It can’t tell you a dark patch is standing water, that a shirt on the ground is soft, or that the thing twenty feet ahead is a person, a deer, or a fence post - distinctions that matter when you’re deciding whether to stop, slow, reroute, or proceed. Camera-based vision that complements LiDAR in unstructured outdoor environments is one of the highest-leverage problems we have. If computer vision is what gets you out of bed, there’s a charter for you here.

We’d be thrilled to hire someone strong on one and curious about the other. We don’t expect both at world-class depth - but we do expect deep ROS2 fluency. That’s the constant.