Starting from the obvious question: "Isn't that the same as a CAE engineer?"
In Japan, CAE engineering is an established profession dedicated to simulation by the finite element method. The Japan Society of Mechanical Engineers certifies competence in the field at Senior Analyst, Grade 1, and Grade 2 levels, and a great many excellent engineers work in this discipline.
Which naturally raises the question: if Mathematical Physics Labo also does FEM analysis, what makes it different from a CAE engineer? I think that question deserves a direct answer. In short, the two are less in competition than working at different stages of the same process.
FEM work divides into before solving, solving, and after solving
In practice, FEM analysis breaks down roughly as follows.
- 1. Turning the phenomenon into equations (formulation): which physics to take as governing equations and which to disregard, and how to couple multiple physical domains such as heat, fluid, and electromagnetics
- 2. Building the model: setting geometry, boundary conditions, material properties, and assumptions
- 3. Solving (discretisation and solution): generating a mesh and computing with a solver
- 4. Assessing the result (validity judgement): deciding whether the numbers obtained are physically correct
- 5. Translating into decisions: carrying the analysis through to design changes and cost judgements
What CAE certification chiefly measures is the ability to run stages 2 to 4 accurately within established standard domains such as solid mechanics, heat, fluids, and vibration. Put differently, it certifies a professional who solves a properly defined problem to a dependable accuracy, and the centre of that value lies in stage 3. This is a highly skilled and genuinely valuable profession.
Mathematical Physics Labo, by contrast, works chiefly in stage 1, the formulation, and in "non-standard" territory where no textbook or off-the-shelf analysis menu exists.
The non-standard territory and the upstream work we take on
Concretely, our position rests on three things.
- Formulating in non-standard territory: airborne molecular contamination (AMC) control in semiconductor cleanrooms, for instance, is a phenomenon in which airflow, chemical transport, and reaction are entangled, and it cannot simply be loaded into an off-the-shelf analysis package. The stage of building up from first principles what belongs in the equations is in fact the hardest part, and the part that determines the outcome.
- Setting the problem upstream: a CAE engineer is in most cases solving a problem that has been handed to them. Mathematical Physics Labo becomes involved from the point of designing what should be solved and how. It is the work of defining the problem before solving it.
- A researcher's background: through particle physics research at Fermilab (Fermi National Accelerator Laboratory) and SLAC (Stanford Linear Accelerator Center), the director of Mathematical Physics Labo spent many years training in building physical models from nothing and verifying computed results against experimental fact.
In a sentence: where a CAE engineer accurately solves a problem that has already been put into solvable form, Mathematical Physics Labo turns the phenomenon into equations that can be solved. Our expertise lies in the stage before the analysis tool is opened.
CAE engineers are collaborators, not competitors
I want to be clear that none of this is intended to place CAE engineers below anyone. If anything, the two roles complement each other. Mathematical Physics Labo formulates the problem from first principles and entrusts a large solution phase to an experienced CAE engineer; or the resulting model is validated from a physics standpoint. Collaboration of this kind arises naturally.
And the very existence of an established CAE certification system is evidence that the market already understands simulation to be worth paying for. Our position sits on that foundation, one step further upstream and in the non-standard territory that lies outside the established menu.
The further AI advances, the more "before solving" is worth
This picture continues directly from the argument in an earlier piece, "Does FEM Consulting Lose Its Value in the Age of AI?". The more generative AI and open-source tools drive down the cost of stage 3, the solving, the more the centre of value shifts toward stage 1, the formulation, and stage 4, the validity judgement: toward what happens before solving and after solving. Mathematical Physics Labo's decision to make the upstream its main ground reflects exactly this shift.
How to start a conversation
"We want to handle a phenomenon that will not load into a standard analysis tool." "We would like to talk from the stage of working out what should be analysed in the first place." Problems whose difficulty lies precisely in the formulation and the framing of the question are what Mathematical Physics Labo is best suited to. If you have a topic involving non-standard physics, including clean air control in cleanrooms, please do get in touch.