The Autonomy Apprenticeship
You hire the junior. We make them productive.
-
Participants can interact with the instructor,
ask questions, seek clarification,
and receive immediate feedback. -
Access to the session that was originally
conducted in real time. -
Cuts future costs by acting as a permanent, searchable database to onboard new hires for free.
-
We will provide you with the notes for the course.
-
Engineers get direct access to the instructor
for ongoing implementation support and Q&A.
Engineered for Your Team. Calibrated to Your Codebase.
Customized to Your Team
Unlike generic public courses, our corporate training can be tailored to your company’s technology stack, workflows, skill gaps, and business objectives.
A structured, gate-based apprenticeship that takes a strong junior engineer and produces a productive autonomy engineer - with evidence that it is working.
Your employee throughout Your payroll, your team, your product, your codebase. We own the training program, not the employment relationship.
Senior PX4 and ROS 2 engineers are scarce and expensive, and the pool is not growing fast enough to meet demand. Every company in this sector is competing for the same few thousand people, and the ones who win are mostly the ones who pay most. That is a bidding war, not a hiring strategy, and it does not get cheaper next year.
There is a second pool nobody is fishing in. Strong junior software engineers, recent graduates with robotics coursework, and career-changers with real programming ability are plentiful and affordable. What stops companies hiring them is not the salary. It is that nobody has the senior bandwidth to bring them up to productivity, and a junior who takes eighteen months to become useful is a cost centre with a long fuse.
The Autonomy Apprenticeship removes the constraint. The apprentice is your employee from day one — your payroll, your team, your product. We are responsible for the training programme and for producing evidence, on a fixed schedule, that it is working. The division is deliberate: employment decisions are yours and should be, and technical judgement about whether someone is becoming an autonomy engineer is ours. Where the answer is no, we will tell you plainly and early rather than at month eleven.
Twelve months, four quarters, one defined capability gate per quarter. An apprentice who does not clear a gate does not simply continue. We tell you, and we tell you what we recommend doing about it.
Rhythm
Structured enough to work, light enough to survive a real project schedule.
- One live training session per week, approximately two hours, cohort-based where you run more than one apprentice
- One technical mentoring and code review session per apprentice per fortnight
- Protected self-directed learning time, approximately four hours per week, which you commit to and protect
- A quarterly written assessment against the capability gate, shared with you and the apprentice
- A quarterly review call with the workplace supervisor
The protected study time is not a formality. It is the one thing the program requires from you that is easy to quietly cancel when a deadline slips, and the one thing whose absence reliably causes an apprentice to miss a gate.
Funding
Federal and state apprenticeship funding mechanisms exist and are actively used for technical apprenticeships, including U.S. Department of Labor program and state-level workforce development schemes. Eligibility and amounts vary by state and by employer, and we do not administer them on your behalf, but we will point you at what applies. Registered apprenticeship status, where you choose to pursue it, is a separate administrative process with its own timeline. We can structure the program to be compatible with registration without requiring it.
Investment
Pricing is quoted per apprentice, with a reduced per-head rate for cohorts of three or more. The program opens with a pilot quarter so that neither side commits twelve months to an untested fit — at the end of Quarter One you have a written assessment in hand and a real basis for deciding whether to continue.
Key Business Outcomes & ROI
The program changes the arithmetic of hiring junior. It compresses the productivity lag from roughly eighteen months to a defined twelve-month schedule with measurable checkpoints, converting an open-ended cost into a budgeted one. It protects senior capacity by removing teaching from your most expensive engineers' calendars — their involvement drops to a quarterly review call and normal code review, rather than daily hand-holding. It de-risks the hire through quarterly written assessment, so a bad fit surfaces in month three instead of month fourteen, when the sunk cost has become an argument for keeping someone who is not going to make it. And it compounds through retention: an apprentice who becomes productive on your product, in your codebase, in your domain, is materially more likely to stay than a senior hire poached at a premium who can be poached again next year. The retention argument is not sentimental. It is the strongest financial case for the program.
Core Programme Objectives
The apprenticeship is structured around one principle: an autonomy engineer is made by working on real hardware and real code under review, not by watching lectures. Apprentices build fluency in the PX4 and ROS 2 stack through high-fidelity Gazebo Software-in-the-Loop environments, where failure injection costs nothing and a mistake destroys no airframe. They develop diagnostic literacy — the ability to interrogate the pxh> console, the EKF2 estimator, and ULog forensic tools rather than searching forums — because the difference between a junior and a productive engineer in this field is mostly the ability to find out why something did not work. They learn the architectural boundary between the real-time flight controller and the Linux companion computer, which is where most integration defects live. And from the second half of the program onward they work inside your repository on your tickets, so the capability they acquire is capability in your system, not in a teaching sandbox. Each quarter closes with a practical, terminal-based gate. Gates are pass or fail, assessed in writing, and shared with both you and the apprentice.
Skills You Will Walk Away With
-
1
Foundations and Environment: bring up, fly, and explain.
- Linux fluency at working level: build systems, toolchains, shell, and the git workflow your team uses
- Build PX4 from source: understand the firmware layout and parameter system
- Stand up a Gazebo SITL environment and fly an autonomous mission end to end
- Map QGroundControl's visual indicators to the underlying pxh> commands and parameters
-
2
Middleware and Offboard Control: write a control node that behaves correctly when the vehicle does not.
- ROS 2 nodes, topics, services, lifecycle, and the QOS settings
- The uXRCE-DDS bridge between the Linux companion computer and the NuttX
- Coordinate frame transformation between PX4's NED convention and ROS 2's ENU
- Offboard control loop programming, setpoint streaming, clock offset, and real-time synchronisation
-
3
Safety and Diagnostics: diagnose a fault nobody warned them about.
- Failsafe architecture: Hold, Return, Land, Terminate and the parameters
- Sensor-in-the-loop failure injection: battery drop, data link loss, geofence breach, GPS degradation, sensor rejection
- EKF2 interrogation: how the vehicle constructs its estimate of where it is, and how that estimate fails
- Flight data forensics with ULog, Flight Review, and PlotJuggler
-
4
Productive on Your Product: ship something your seniors would have shipped.
- Full-time work inside your repository, on your backlog, against your review standards
- Ownership of scoped feature tickets from specification through integration test
- Participation in code review as a reviewer, not only as a reviewee
- Integration and validation work on your airframe and your test rig
Questions? Check here
Yes. All live sessions are recorded and the archive is yours permanently, which means the material becomes an internal onboarding resource for future hires at no additional cost.
Strong general programming ability, working Linux comfort, and intermediate Python. Robotics coursework helps but is not required; C++ exposure helps but can be built. We are happy to review a shortlist or sit in on technical interviews before you commit to a hire — we would rather help you pick well than train someone who was never going to fit.
One works. Two to four works better and costs less per head, because the weekly session becomes a cohort session and apprentices learn a great deal from watching each other's code get reviewed. Above six we would split into two cohorts.
The technical spine — flight controller architecture, the ROS 2 boundary, failure handling, log forensics — transfers. The specific exercises are rebuilt around what you actually run. We do the same for ArduPilot-based and hybrid stacks.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
No. The program runs perfectly well as a private arrangement. We structure it to be compatible with registered apprenticeship standards so that you can pursue registration if you want the funding or the credential, but registration is your administrative process on your timeline, not a precondition.
Yours, throughout. Your payroll, your benefits, your team, your product. We are a training provider, not a staffing agency, and we take no position on compensation, promotion, or termination. Those are your decisions and they should be.
A quarterly review call with the workplace supervisor, plus normal code review on the apprentice's work from Quarter Three onward. Teaching, curriculum, and one-to-one technical mentoring are ours. If you find your seniors being pulled into daily instruction, the program is not working as designed and we want to know.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
Yes. All live sessions are recorded and the archive is yours permanently, which means the material becomes an internal onboarding resource for future hires at no additional cost.
Strong general programming ability, working Linux comfort, and intermediate Python. Robotics coursework helps but is not required; C++ exposure helps but can be built. We are happy to review a shortlist or sit in on technical interviews before you commit to a hire — we would rather help you pick well than train someone who was never going to fit.
One works. Two to four works better and costs less per head, because the weekly session becomes a cohort session and apprentices learn a great deal from watching each other's code get reviewed. Above six we would split into two cohorts.
The technical spine — flight controller architecture, the ROS 2 boundary, failure handling, log forensics — transfers. The specific exercises are rebuilt around what you actually run. We do the same for ArduPilot-based and hybrid stacks.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
No. The program runs perfectly well as a private arrangement. We structure it to be compatible with registered apprenticeship standards so that you can pursue registration if you want the funding or the credential, but registration is your administrative process on your timeline, not a precondition.
Yours, throughout. Your payroll, your benefits, your team, your product. We are a training provider, not a staffing agency, and we take no position on compensation, promotion, or termination. Those are your decisions and they should be.
A quarterly review call with the workplace supervisor, plus normal code review on the apprentice's work from Quarter Three onward. Teaching, curriculum, and one-to-one technical mentoring are ours. If you find your seniors being pulled into daily instruction, the program is not working as designed and we want to know.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
One works. Two to four works better and costs less per head, because the weekly session becomes a cohort session and apprentices learn a great deal from watching each other's code get reviewed. Above six we would split into two cohorts.
The technical spine — flight controller architecture, the ROS 2 boundary, failure handling, log forensics — transfers. The specific exercises are rebuilt around what you actually run. We do the same for ArduPilot-based and hybrid stacks.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
No. The program runs perfectly well as a private arrangement. We structure it to be compatible with registered apprenticeship standards so that you can pursue registration if you want the funding or the credential, but registration is your administrative process on your timeline, not a precondition.
Yours, throughout. Your payroll, your benefits, your team, your product. We are a training provider, not a staffing agency, and we take no position on compensation, promotion, or termination. Those are your decisions and they should be.
A quarterly review call with the workplace supervisor, plus normal code review on the apprentice's work from Quarter Three onward. Teaching, curriculum, and one-to-one technical mentoring are ours. If you find your seniors being pulled into daily instruction, the program is not working as designed and we want to know.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
No. The program runs perfectly well as a private arrangement. We structure it to be compatible with registered apprenticeship standards so that you can pursue registration if you want the funding or the credential, but registration is your administrative process on your timeline, not a precondition.
Yours, throughout. Your payroll, your benefits, your team, your product. We are a training provider, not a staffing agency, and we take no position on compensation, promotion, or termination. Those are your decisions and they should be.
A quarterly review call with the workplace supervisor, plus normal code review on the apprentice's work from Quarter Three onward. Teaching, curriculum, and one-to-one technical mentoring are ours. If you find your seniors being pulled into daily instruction, the program is not working as designed and we want to know.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
Yours, throughout. Your payroll, your benefits, your team, your product. We are a training provider, not a staffing agency, and we take no position on compensation, promotion, or termination. Those are your decisions and they should be.
A quarterly review call with the workplace supervisor, plus normal code review on the apprentice's work from Quarter Three onward. Teaching, curriculum, and one-to-one technical mentoring are ours. If you find your seniors being pulled into daily instruction, the program is not working as designed and we want to know.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.
We tell you, in writing, at the gate — and if it is obvious sooner, sooner than that. The assessment states what was missed, whether we believe it is recoverable, and what we recommend: remediation, an extended quarter, or ending the programme. The pilot quarter exists precisely so that this conversation, if it needs to happen, happens in month three rather than month fourteen.