PROCUREMENT QUESTIONS

The questions a buyer asks before the first invoice

Contracting, intellectual property, assessment, export control, insurance and data handling. Answered here so you do not have to spend a call extracting them.

If your question is not here, ask it. We would rather answer it once in public than twenty times in private, and anything that comes up more than twice gets added to this page.

Contracting and intellectual property
Who owns the code you write for us?

You do. Work produced specifically for your programme is yours on payment, including source, build scripts and test artefacts.

What stays ours is background IP: the reference implementations, our internal libraries, templates and process assets that existed before your engagement. You get a licence to use those inside the delivered product, which is what makes the delivery work; you do not get the right to resell them as a component.

We write this split into the contract explicitly rather than leaving it to a general clause, because "foreground and background IP" means different things at different OEMs.

Do you sign a Development Interface Agreement?

Yes, and we prefer to. A DIA names which safety work products each side produces, who reviews them, who approves them, and which interfaces carry a joint review.

If you have a DIA template, we work to it. If you do not, we bring one and you edit it. Either way it is signed before design work starts, not retrofitted before the assessment.

Will you work as a subcontractor to our Tier 1?

Yes. In that arrangement the DIA is usually between us and the Tier 1, with the OEM requirements flowed down. We will ask to see the flow-down rather than assume it, because a requirement that never arrived is the most common cause of a late finding.

Do you use subcontractors, and will we know?

For specialist work such as PCB layout, RF, power electronics and mechanical design we bring in engineers we have worked with before. You are told who, for which scope, before they start.

Certification-critical work stays with us. We do not subcontract the safety case for work we did not do.

Safety, assessment and standards
Do you produce our safety case, or support ours?

Both are normal engagements and they are priced differently.

  • We produce it: we write the hazard analysis, derive the concepts, assemble the argument and the evidence, and answer for it.
  • We support yours: your team owns the case and we fill named gaps, typically FMEDA, the software safety lifecycle, or the traceability that ties them together.

What we do not do is put a signature under an argument we have not been allowed to inspect.

Will you attend the assessment?

Yes, for the scope we produced. We sit in the interviews, answer the assessor’s questions on our work products, and take the findings that belong to us.

Attendance is quoted as work rather than bundled invisibly, because assessment weeks are long and pretending otherwise inflates the day rate for everyone else.

Which standards do you actually work to?
  • ISO 26262:2018 parts 2 to 9, road vehicles
  • IEC 61508 for industrial items outside the automotive scope
  • ISO/SAE 21434:2021 and UN R155 / R156, cybersecurity and software update management
  • ISO 21448 (SOTIF) for perception and control behaviour
  • Automotive SPICE for process assessment support
  • MISRA C:2023 and MISRA C++, AUTOSAR C++14 where the platform requires it

Terms used above are defined on the glossary page.

Are your reference implementations certified?

No, and no component can be. They are developed as SEooC: certification applies to an item in a vehicle, not to a library.

What ships with a licensed component is a safety manual stating the assumptions of use, the integration constraints, and the claims the component makes about itself. Your item-level analysis decides whether those claims hold in your system. See licensing.

Is your platform a qualified tool?

No. Nothing we publish is a qualified software tool under ISO 26262-8 clause 11, and it does not substitute for one. It is assistive: every output is reviewed through your existing qualified process.

If your programme needs a tool confidence argument for something we supply, that is a scoped piece of work and we will say so rather than hand you a claim you cannot defend.

Confidentiality, data and export control
What happens to data we send you?

Programme material arrives under NDA and is held only for the engagement. We do not put client material into third-party services for processing.

The demonstration platform on this site is deliberately built so that the question does not arise: files you load into it are parsed in your own browser and never leave your machine. That is why the FMEDA, SBOM, coverage and requirements importers all run client-side.

Are we subject to export control?

Possibly, and it is worth establishing early rather than at delivery. We are a United States entity, so US EAR applies to what we hand over; ITAR applies if the item is defence-listed. EU dual-use rules can apply to your side of the same transfer.

We ask about the end use and the destination before quoting, and if a licence is required we say so instead of shipping and hoping.

Do you carry professional indemnity insurance?

Yes. Certificate and limits are provided during supplier qualification, along with the credential verification links for the certifications listed on the about page.

Liability caps are negotiated per engagement and we do not pretend the default is unlimited.

Can you work in our language?

Programmes run in English. Written deliverables are English by default; German and Japanese review support is available where the assessment is conducted in those languages.

Standard vocabulary stays in English in every language, as it does in the standards themselves. ASIL, SPFM, PMHF and FTTI are not translated.

Engagement shape
How small an engagement do you take?

A scoped work package is the normal starting point: one HARA, one FMEDA, one TARA, one bring-up. Those are typically two to six weeks and they exist so both sides can see how the other works before committing to a programme.

We do not take open-ended staff augmentation. It produces bad safety work, because nobody owns the argument.

Which silicon have you worked on?

Programmes to date have run on Renesas, NXP (including Freescale-era parts) and Texas Instruments automotive silicon, across microcontrollers and the larger application processors.

The published reference implementations name the part each one's timing was measured on, because a latency claim has to be measured somewhere. That is a statement about where the evidence came from rather than a supported-hardware list: the components are written against an abstraction layer and two of them carry no target dependency at all.

What matters more than the part number is the class of architecture: a lockstep real-time MCU, a motor-control MCU with gate-driver and resolver interfaces, a zonal or gateway controller with a hardware security module, an application processor sitting beside a real-time core, or a TSN-capable network processor. Those are set out with their evidence on the licensing page.

Tell us the part your programme uses. If we have not shipped on it we will say so, and we will tell you what the port would involve rather than quoting for a discovery phase to find out.

Which tools do you work in?

We adapt to the programme rather than requiring our own stack.

  • Requirements and traceability: ReqIF interchange with Polarion, codebeamer or DOORS
  • Safety analysis: medini analyze, or spreadsheet-based FMEDA where that is what the programme uses
  • Static analysis and coverage: Polyspace, Helix QAC, LDRA, VectorCAST, gcov / llvm-cov
  • Embedded toolchains: IAR, Green Hills, GCC, Lauterbach for debug
  • Hardware: Altium and KiCad for capture and layout

If your toolchain is not on this list, say which one and we will tell you honestly whether we know it.

Where are you, and which hours?

norxs Technology LLC is registered in Wyoming, United States. Engineering runs to European and Asian programme hours as well as North American, and review meetings are scheduled to the programme rather than to us.

Every engagement runs under NDA, which is why the work on this site is described without client names. Security reporting and data handling detail sits on the trust page.

Question that is not answered here?

Send it. If it is a fair question for a supplier, it belongs on this page.

Ask directly Licensing terms