AMR vs humanoid jobs: which robotics career to choose

If you are choosing a robotics path in 2026, the loud options are warehouse AMRs and humanoid labs. Both hire. They hire different people, in different cities, with different tolerance for research risk. This is a practical comparison so you can pick a stack and a first job instead of collecting every framework.

Start by browsing live AMR jobs and humanoid jobs. Then use the skill deltas below.

AMR vs humanoid at a glance

AMR programs live in warehouses and factories. Success is a robot that moves pallets on Tuesday without a babysitter. You will talk about maps, traffic, docks, WiFi, and the 3PL customer who changed the aisle layout.

Humanoid programs live in labs that want a general-purpose machine in spaces built for people. Success is a policy or controller that does not fall, plus a data engine that does not starve. You will talk about whole-body control, teleoperation, GPU training, and hands.

Neither is "more real." AMRs have more customer floors. Humanoids have more research prestige and more opaque hiring bars.

What AMR robotics work feels like

AMR software still matters. So does standing next to a charging stand at 6 a.m. Typical titles: navigation engineer, fleet software, deployment engineer, test, applications, and site reliability for robots.

The stack is often ROS 2, C++, Linux, and a vendor or in-house fleet layer. PLC shows up when the cell talks to the rest of the warehouse. See ROS 2 jobs and PLC robotics jobs.

If you like customer contact and shipping criteria, this is the path. Entry level robotics jobs look healthier here than in humanoid labs.

What humanoid robotics work feels like

Many jobs are on site five days a week in a small set of cities. Software bars are high for perception, controls, and learning. Operator and data-collection jobs exist if you want in without owning the policy.

The stack may not be ROS. CUDA, PyTorch, whole-body control, and proprietary middleware show up instead. Compare PyTorch jobs with humanoid jobs and company pages such as Figure and Apptronik.

If you want to bet on a lab, live in one of those cities, and accept that the product timeline is someone else's fundraising cycle, this is the path.

Skills that actually transfer

Shared core: C++, Python, Linux, sensors, and the humility to debug a system you cannot fully simulate.

AMR extras: ROS 2 navigation, fleet software, industrial networking, sometimes PLC. Deployment engineers get hired without a SLAM paper.

Humanoid extras: controls at the body level, CUDA, large-scale learning, and comfort with stacks that are not ROS. Operators and data jobs exist if you want in without owning the policy.

If you already know ROS 2, AMRs will use it tomorrow. Humanoids might not. If you already know PyTorch well and have never touched a motor driver, humanoid perception teams are a better fit than an AMR commissioning job.

Hiring volume, seniority, and pay

AMR hiring is broader at junior and mid. Integrators, test, and field software show up for people coming from ME, EE, or CS. Humanoid hiring is barbell: operators and data workers on one side, senior researchers and staff engineers on the other.

Compensation follows. Humanoid labs in the Bay Area post aggressive software bands. AMR deployment in secondary cities can pay less cash and still be a better first job if you want hardware hours. Filter senior and lead listings only after you know which loop you want.

How to choose a robotics career path

Choose AMRs if you want customer contact, shipping criteria, and a portfolio of sites. Choose humanoids if you want to bet on a lab, live in one of those cities, and accept slower product proof.

You can switch later. Navigation, controls, and Linux travel. The thing that does not travel is "I only trained models and never owned an on-robot crash."

Build one project in the domain you want: a Nav2 stack for AMRs, or a sim locomotion or manipulation demo for humanoids. Then apply through the job list instead of spraying every "robotics" title on LinkedIn.

For the longer on-ramp, read how to become a robotics engineer. For what postings emphasize this quarter, read skills in live job posts.