Full-Time

Staff Perception Hardware Engineer

Updated on 9/4/2026

Apptronik

Apptronik

201-500 employees

Develops human-centered robotic systems and components

Compensation Overview

$240k - $288k/yr

Sunnyvale, CA, USA

In Person

Master's, PhD

Category
Hardware Engineering (1)
Required Skills
Python
Machine Learning
C/C++
Robotics
FEM/FEA

Get referred to Apptronik

See people who can refer or advise you

Requirements
  • A proven track record of taking complex perception hardware, including custom-defined sensor modules, from the breadboard phase through Design Verification Testing and into mass production.
  • Proven experience working directly with optical vendors, CMOS manufacturers, and contract manufacturers to define, build, and test custom camera or sensor modules.
  • Expert-level proficiency with high-end lab equipment, including high-bandwidth oscilloscopes, logic analyzers, and time-domain reflectometers.
  • Deep practical understanding of CMOS sensor tuning, optical lens artifacts such as flare, ghosting, and distortion, and physical mitigation within tight mechanical enclosures.
  • Mastery of the linear algebra and 3D geometry underlying multi-camera arrays, stereo baselines, and visual-inertial sensor fusion, with the ability to write Python or C++ scripts to validate these models on real hardware.
  • Ability to lead technical initiatives and align cross-functional stakeholders without direct reporting authority, using empirical data from the lab to drive decisions.
  • Ability to thrive in a fast-paced, high-growth environment and determine when a design is robust enough for real-world deployment.
  • A master's or Ph.D. degree in Electrical Engineering, Optical Engineering, Computer Engineering, or a related field.
  • At least 8 years of hands-on experience integrating and shipping complex vision hardware systems for robotics, autonomous vehicles, or advanced augmented reality/virtual reality platforms.
  • Ability to sit for prolonged periods while working at a desk and computer.
  • Ability to lift 15 pounds at times.
  • Vision to read printed materials and a computer screen.
  • Hearing and speech sufficient to communicate.
Responsibilities
  • Lead the physical integration of the robot's visual perception stack and drive prototype designs into validated, production-ready camera assemblies meeting field-of-view, thermal, and mechanical-stability targets.
  • Define electrical, optical, and mechanical specifications for custom sensor modules when off-the-shelf components are insufficient.
  • Drive external optical vendors and contract manufacturers through the custom sensor-module build process and ensure adherence to company standards.
  • Architect and own bring-up, Design Verification Testing, and End-of-Line test plans for custom-built sensor modules.
  • Perform hands-on image signal processor tuning, including auto-exposure, white balance, and tone mapping, in real-world environments.
  • Ensure camera data meets machine-learning model requirements and the low-latency needs of human teleoperators.
  • Lead electrical bring-up and validation of the multi-gigabit video backbone, including GMSL2/3 and FPD-Link.
  • Resolve system-level electromagnetic interference and electromagnetic compatibility issues using high-speed oscilloscopes and analyzers to identify frame drops and signal-integrity problems across dynamic robot joints.
  • Implement and measure hardware-level synchronization across cameras, LiDAR, and inertial measurement units, proving sub-microsecond timing accuracy on the physical robot.
  • Design, build, and deploy physical calibration fixtures and calibration algorithms for optical-system alignment.
  • Troubleshoot and resolve intrinsic and extrinsic calibration drift so the calibration process scales from the research laboratory to the factory floor.
  • Serve as the technical adjudicator between Industrial Design, Mechanical, and Thermal teams, using physical testing and prototyping to resolve conflicts involving surface aesthetics, sensor field-of-view constraints, and thermal dissipation.
  • Lead complex, cross-functional problem-solving efforts to resolve critical line-down and launch-blocking hardware bugs.
  • Mentor senior and junior engineers in the lab and establish rigorous physical-testing and root-cause-analysis practices.

Apptronik builds and sells human-centered robotic systems, components, and engineering services. Its products range from robotic components for hobbyists to advanced robotic systems for researchers and industries, driven by mechatronics and control software used to power legged robots like Valkyrie and Hume. These robots are designed to operate in real human environments, enabling safer, more productive interactions by being lighter, safer, and more efficient. What sets Apptronik apart is its practical, user-friendly focus on integrating robots into everyday work and life, not just advancing technology, and its end-to-end approach that combines hardware, software, and engineering services across a broad customer base. The company’s goal is to develop robots that can live and work alongside humans, improving safety, productivity, and quality of life.

Company Size

201-500

Company Stage

Private

Total Funding

$949.8M

Headquarters

Austin, Texas

Founded

2015

Get referred to Apptronik

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • February 2026 brought $520 million, lifting total Series A past $935 million.
  • June 30, 2026 Robot Park expansion scales training from ten robots to hundreds.
  • Apptronik added Waymo, Boston Dynamics, and Amazon veterans in April 2026, strengthening execution.

What critics are saying

  • Figure AI and Agility Robotics already deploy humanoids; Apptronik risks being outexecuted by 2027.
  • Apollo 2 remains a data-collection workhorse; delayed Apollo 3 commercialization traps capital.
  • Safety standards for bipedal humanoids still lag; one serious factory incident could freeze deployments.

What makes Apptronik unique

  • Apptronik pairs Apollo 2 hardware with Google DeepMind's Gemini Robotics data flywheel.
  • Robot Park gives Apptronik unmatched real-world humanoid training volume versus lab-only rivals.
  • Mercedes-Benz, GXO, and Jabil pilots prove Apptronik targets industrial workflows, not toy demos.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Flexible Work Hours

Unlimited Paid Time Off

401(k) Retirement Plan

Company Equity

Growth & Insights and Company News

Headcount

6 month growth

1%

1 year growth

0%

2 year growth

0%
Aegis Capital Corp
Jul 20th, 2026
Aegis Capital Corp. acted as Placement Agent in connection with a ~$2.0 Million Secondary Transaction for Apptronik.

Aegis Capital Corp. acted as Placement Agent in connection with a ~$2.0 Million Secondary Transaction for Apptronik. 7/2026 - Aegis Capital Corp. acted as Placement Agent in connection with a ~$2.0 Million Secondary Transaction for Apptronik About Apptronik Apptronik is a spinout from the Human Centered Robotics Lab at the University of Texas at Austin. Aegis Capital Corporation were founded in early 2016 with the goal of bringing forth the next generation of robots that will change the way Aegis Capital Corporation live and work. Aegis Capital Corporation has built some of the most diverse robots, spanning from exoskeletons to humanoid torsos, biped mobility platforms, and unique robotic arms that lift more than they weigh. The experience and learning from all this work led to the development of Apollo, the world's most advanced humanoid robot. About Aegis Capital Corp. Aegis Capital Corporation ("Aegis") has been in the wealth management and investment banking business since 1984. Aegis is dedicated to providing corporate finance, strategic advisory and related services to public and private companies across multiple sectors and regions. Aegis also provides research and sales and trading services to institutional investors. Aegis offers its investment representatives a conflict free service platform and is able to provide a full-range of products and services including investment banking, wealth management, insurance, retirement planning, structured products, private equity, alternatives, equity research, fixed income and special purpose vehicles.

Semafor
Jul 2nd, 2026
Apptronik opens massive Robot Park to train hundreds of humanoids for real-world tasks

Apptronik has opened Robot Park, a warehouse in Austin, Texas, the size of two football fields where hundreds of its humanoid robots will practice real-world tasks. The facility addresses a critical industry challenge: the lack of real-world data needed to improve robot AI models. The new testing site can accommodate hundreds of robots, compared to roughly 10 at Apptronik's previous facility. The humanoids will practise tasks for manufacturing customers, including packaging, sorting tools and moving boxes. Whilst companies use digital simulations for training, these cannot account for real-world factors like hardware ageing or slipping on surfaces. CEO Jeff Cardenas said the facility allows the company to "capture that physical nuance and adapt quickly." Apptronik will share data with research partner and investor Google DeepMind, which will integrate findings into Gemini Robots, an AI model used across the robotics industry.

Yahoo
Jul 2nd, 2026
Apptronik turns robot simulations into reality.

Apptronik turns robot simulations into reality. Rachyl Jones Thu, July 2, 2026 at 2:24 PM PDT The news. Humanoid-maker Apptronik has opened Robot Park, a factory in Austin, Texas, where its robots will practice real-world tasks across a warehouse the size of two football fields. One of the biggest hurdles for robot companies is a lack of real-world data: There aren't enough robots in the field to significantly improve the models running them, and existing testing sites can only hold a small number. Apptronik's new warehouse will see "hundreds" of the company's latest humanoids in operation, spokesperson Liz Clinkenbeard told Semafor. Its previous testing site could only hold roughly 10 at a time. The bots will practice tasks they'll perform for customers - often manufacturers - like packaging items together, sorting tools, and moving boxes around shelves. Along with real-world tests to train robot systems, companies also use digital simulations, where renderings of the robots practice tasks in the digital world. Those simulations can run all day, every day, but they don't account for limitations like aging hardware or real-world occurrences like a robot's foot slipping on the ground. By testing at Robot Park, the company can "capture that physical nuance and can adapt quickly," CEO Jeff Cardenas said. Apptronik will share the data it gets with Google DeepMind, a research partner and investor. Google can then integrate that data back into Gemini Robots, the AI model used across the robot industry. Know more. As part of Apptronik's Tuesday announcement of Robot Park, it also unveiled the latest line of its Apollo humanoid robots. Notably, it upgraded its wheeled robot, which looks like a humanoid mounted to a wheeled base rather than legs, because of high demand, Clinkenbeard said. The industry has not yet determined adequate safety standards for two-legged robots to operate alongside humans. But wheeled robots are currently allowed.

eWeek
Jul 2nd, 2026
Apptronik debuts new Apollo 2 humanoid, launches 90,000-square-foot 'Robot Park'

Apptronik debuts new Apollo 2 humanoid, launches 90,000-square-foot 'Robot Park' Image: Apptronik Jul 2, 2026 eWeek content and product recommendations are editorially independent. eWEEK may make money when you click on links to its partners. Learn More Apptronik wants to solve one of robotics' biggest bottlenecks: collecting enough real-world experience to make humanoid robots genuinely useful. The Austin startup has opened a roughly 90,000-square-foot Robot Park where fleets of humanoid robots perform industrial work while continuously generating data used to train AI models. The company says the facility serves as a production-scale learning environment rather than a traditional research lab. The announcement reflects a broader shift across the robotics industry. Companies are increasingly competing not just on robot hardware, but on who can gather the most valuable real-world data to improve AI systems. Apollo 2 steps into the spotlight. Alongside the facility, Apptronik unveiled Apollo 2, its latest humanoid robot platform. It comes in two configurations, bipedal and wheeled, designed to handle different operational needs. The wheeled version targets industrial settings where stability and efficiency matter most, while the bipedal version is built for environments closer to human spaces, where walking ability and adaptability are required. According to Apptronik, Apollo 2 has been operating as a data-collection workhorse for over a year already. The system is now deployed across Robot Park sites and at partner and customer facilities to generate training data for future models. A key pillar of Apptronik's strategy is its partnership with Google DeepMind. Data gathered from Apollo 2 is used to advance Gemini Robotics, DeepMind's foundational AI models for robotics. Apptronik CEO Jeff Cardenas described the system as a feedback loop between real-world operations and AI training. The company has built what Cardenas called "a factory that produces robots, we also have a factory that produces data," according to Reuters. He added that Robot Park serves as the engine behind production-grade AI development, linking physical robots directly to model training at scale. Robots train for real work. The robotics industry has long struggled with a core problem: robots can perform impressive demos but often fail in real workplaces. Apptronik's approach aims to close that gap by prioritizing continuous exposure to real environments - logistics, manufacturing, and retail operations - rather than controlled lab demonstrations. If successful, this model could shorten the gap between prototype robots and commercially useful machines that actually perform repeatable work at scale. What Apptronik is really building. At its core, Apptronik is not just building robots; it is building a data pipeline for embodied AI. Robot Park functions like a manufacturing line, but instead of producing hardware, it produces training data. That data is then used to refine AI models that control future generations of robots. The partnership with Google DeepMind is especially significant. If Gemini Robotics becomes a widely used foundation model, Apptronik could benefit from early access and integration, but it also raises a long-term question: whether the intelligence it helps train becomes broadly shared across competitors. The company is also clearly separating roles between its current and future hardware. Apollo 2 is positioned as a workhorse for data collection, while a future system, Apollo 3, is expected to be the first fully commercial, scaled product. Also read: Japan wants 10 million AI-powered robots operating by 2040 as it expands robotics across manufacturing, caregiving, infrastructure, and disaster response. Aminu Abdullahi is a B2C and B2B technology and finance writer with more than six years of experience covering enterprise IT, cybersecurity, cloud computing, artificial intelligence, fintech, business software, and emerging technologies. His work has appeared in publications including TechRepublic, eWEEK, Channel Insider, Geekflare, Enterprise Networking Planet, eSecurity Planet, CIO Insight, and Webopedia. With a technical background in computer science, he specializes in translating complex technology topics into clear, accessible content for business leaders and decision-makers.

Business Insider
Jun 30th, 2026
Texas robotics startup opens 'Robot Park' to train $5.5B humanoid workforce

Apptronik, a Texas-based humanoid robotics startup valued at over $5.5 billion, has opened a 90,000-square-foot "Robot Park" in Austin where its Apollo robots practice tasks like loading boxes and sorting items. The facility operates seven days a week, generating training data to improve the AI models that control the robots. The company, which spun out from the University of Texas in 2016, has raised approximately $1 billion. Google's DeepMind and Mercedes are among its investors, with Mercedes deploying Apollo robots in factories for assembly-line tasks. Apptronik's current Apollo 2 model stands 6 feet tall, runs for four hours, and can lift 55 pounds. The company is developing both wheeled and legged versions, with wheeled robots expected to deploy sooner due to safety considerations. Apptronik plans to establish additional Robot Parks globally.