Introducing Senslogic and WaveMe: Closing the Gap Between Optical Design and Reality

This tech talk is about Senslogic and the philosophy behind it. After over two decades in optics, one realization stands out, which is that there is a disconnect between theory and practice. Maybe the joke that “it works in theory” has some merit. Obviously, this is not the case everywhere, but it is so frequently enough.

We have design tools with good predictive capabilities, diffraction theory works, but too often it stops there, and the system is aligned using small apertures and business cards. Once all elements are in place, the system gets to see “first light” and then starts the slow march toward reaching the expected, or should I rather say “desired” performance because when millimeter-sized apertures and business cards are your tools, hitting target performance becomes entirely unexpected. Hence, what follows is, if we allow ourselves a euphemism, tweaking.

Enough Tweaking

Why do we tweak? Well, obviously because we want a different result. Sometimes we even tweak because perfection is out of reach and we are left between choosing the lesser of two evils. But if there is a time and place for tweaking, that time is during the design phase. If we cannot have it all, this is where we should make the choice. Not on the production floor, not by someone without the necessary information to make that choice, especially not after sinking all those resources into reaching that stage of the product.

As mentioned earlier, the predictive power of the tools available to us is really good. There really is no excuse for being ignorant of how an optical system works. When needed, we should incorporate techniques like the Monte Carlo or Latin Hypercube Sampling to gain a comprehensive view of all variables affecting system performance. These techniques allow us to grasp a much larger set of variables, ensuring that design decisions are driven by data rather than assumptions.

When it comes to tweaking, this is where it truly belongs. It’s at this stage—where all variables are visible and data-driven trade-offs can be made—that tweaking should occur.

Build the Design by Designing the Assembly Process

Please allow a short segue for context. Tennis has a curious scoring system, rooted in the idea of a clock and coming full circle. In this tech talk, we’re at ‘30’—well on our way, but not yet at game point. Nevertheless, it is a huge point.

Each component of an optical system starts as a discrete element—waiting to find its precise place within the final assembly, and if we are responsible for finding it, we need tools. Enter the WaveMe Toolbox, designed specifically to ensure precision in building optical systems with discrete elements.

BeamNotes

The simplest of the toolboxes provided by WaveMe is the BeamNotes tool. Yet, it is probably the one that best captures the spirit of this tech talk. The measurement it provides isn’t just simple—it’s deliberately straightforward, focusing on a basic centroid measurement enhanced with a few key tools to remember positions, easily spot that centering is according to tolerances, and pass notes between design and assembly teams.

BeamNotes helps us achieve precise optical alignment along a laser beam, even when the mechanical setup is less than ideal. It is a crucial element in the assembly workflow, enabling seamless communication and precise alignment to ensure that the right information is available exactly when needed.

Shack-Hartmann

I get it, the BeamNotes tool is not impressive. However, the Shack-Hartmann tool is. Maybe you already use a SHS (Shack-Hartmann Sensor). How long does it take to get a result? Do you have to align the sensor? Aligning a sensor within a system that’s still being adjusted creates a Catch-22—how can you align one part of the setup while other parts are still moving? The Shack-Hartmann toolbox in WaveMe elegantly sidesteps this issue with its synthetic reference calibration.

Does your sensor require you to define regions for the spots? Most of the tools out there are way too complex for practical assembly work. The time to obtain a single result is in the minute range, or more. This I personally find unacceptable. With WaveMe’s Shack-Hartmann toolbox, making a wavefront measurement is as fast as selecting the tool from the menu—often, it’s just a matter of turning on the program. If I had the chutzpa of Amazon, I’d call it ‘1-click measurement.’ But since you have to select the tool and enable it, let’s call it ‘2-clicks to perfection’. That’s it, there is nothing more to do, and there shouldn’t be. No regions of interest or exposure settings. All automated. You can read about it in the tech talk about calibration. It’s pretty neat.

Phase-shifting Interferometry

Element centering and collimation are crucial in optical assembly, but optical systems are more complex than that. Lenses can be squeezed by mounts, mirrors bent by coatings or mechanics, and lenses can be inserted in the wrong direction. It happens. Earlier, I did try to make the analogy with a game of tennis and the clock metaphor, and this is the process coming full circle. We started with models, found the requirements and designed optics. We laid out a strategy for assembly and provided the tools. But since this is optics, and despite our best efforts there is a final step to be made, a no excuses, no if or buts. Just a pure – this is what it is. This is the final grade for our efforts.

Do the results agree with the models? Do the assembly instructions give us the same system each time, no matter who builds them. A phase-shifting interferometer will give us the answer. The phase-shifting interferometer in WaveMe is not only accurate, it’s also fast. With a USB3 camera and a fast actuator, 15 wavefronts per second is entirely possible. It is also quite versatile since it assumes that the user may want to alter the system under test in some way, like an OMEMS or deformable mirror. In order to reach speeds that are limited only by the transport interface of the camera, the phase-shifting tool has to transition between synchronous and asynchronous states so that the different physical parts can move when the interferometer doesn’t need them to be at some fixed state.

A point diffraction interferometer is almost always a custom tool but it is not difficult to build. For high-end imaging systems, like lithography, it will always tell you what you need to know.

Full Circle & Bonus

We have now come full circle. We have the results to feed back to our models or the way we assembly our systems. Imagine having done this a couple of times, tuned the process, learned from the mistakes. After a few turns of this wheel, we will gain enough confidence that already after the modelling process, we will know the product we will have two or three years down the road. The value in that may vary but this is what the learning process provides, but until we have the tools for it, this loop cannot be closed. In the tennis analogy, our game is done, but until the match is won, there is the next game and if we don’t bring the experiences from the last game to the next, we are likely to repeat whatever mistakes we did then, and our opponent will thank us for it. Let’s not do life too simple for the competition.

A few work before we close this tech talk. The field of optics is huge. It’s difficult to provide one tool for everything, but WaveMe is a flexible tool. The program itself is not much more than a pipeline, a compositor and a modules interface. Modules bring their own user interface and can interact with other modules (or tools). The pipeline and the module interface are an open API. If a tool does not exactly do what you like, it’s not difficult to add one to the pipeline to get the result you are looking for. Anyone can write a WaveMe module and interact with the messages in the pipeline, and to streamline this process, WaveMe comes with open-source templates that can be copied and extended in any way imaginable. This is in fact how all the currently available tools were once developed.