“Winning the Innovation Award created great awareness”

Did winning the award have a noticeable impact on your company?
Antolovic: Absolutely! It certainly created more awareness for our company, brought more people to stop by to our booth at LASER and marked a great milestone for us. In July we also announced the complete acquisition by Zeiss, with whom we had been working closely for a long time—yet another milestone! All of this makes our team proud and reiterates of course our commitment.

You won the award with a special camera. What are its advantages?
Antolovic: The SPAD Alpha was fully built in-house. It is based on our key technology: Single Photon Avalanche Diodes, which transform light directly into digital signals. It counts photons by sensing every photon as a particle of light. Due to this direct digitization, we can operate the camera extremely fast, and we can also detect very weak light. These are the key advantages: it's extremely fast and really good for low-light applications.

Does direct transformation from photon to digital bits mean that light intensity doesn’t need to be strong, because there is nothing like a weak bit?
Antolovic: Exactly. Light is quantized. It is both a wave and a particle. We use the property of light as particles. We count them and can detect approximately fifty percent of the photons. So, every second photon is counted. But what’s important: we can reliably distinguish whether it’s a photon or not. There is no more ambiguity as known from analog cameras.

Could you name some key customer benefits?
Antolovic: Our approach results in better image quality and the ability to extract information that is not possible to extract with other detectors. This ranges from the fact that we offer the highest photon fidelity, which is important for quantum applications, to the fact that we are extremely fast—which means that we can extract very fast phenomena that are not visible by using standard cameras. For example, in cancer sectioning we know that brain tumors have different spectra compared to healthy tissue. But it isn't only the spectrum, but also the timing fingerprint of the photons emitted, which is different. With our cameras, we can extract this information and thus determine more accurately and precisely where healthy tissue is located and where cancerous tissue is located.

You offer an additional method for characterizing tissue by measuring the specific patterns and frequencies with which tissue emit photons after excitation?
Antolovic: Yes, it is similar to radioactive decay. The fluorescent molecules have a certain timing fingerprint that our very fast and precise technology can detect. Basically, in addition to intensity and color, there is also the timing aspect which we use for more precise diagnoses.

Could you please elaborate on SPAD technology?
Antolovic: It is very important for our company to build SPADs using standard semiconductor technologies. Therefore, it is very scalable, meaning that we can build anything from a single pixel to one megapixel or more. There are no fundamental limits or limitations in the resolution. Before we started, most of the photon counting companies had only one pixel. Instead, we focused on SPAD arrays. There were a few companies that created smaller SPAD arrays before us, but we were first to really commercialize a high resolution SPAD-camera in 2021, which was a 512 x 512-pixel solution. We build both image sensors and cameras, so we are a full supply chain company. Our key competence is the design of SPAD-image sensors. This is our biggest advantage, and that's also how we position due to the very high entrance barrier in the semiconductor market. We understood that many niche markets are under-commercialized. We address these markets and applications which are outside of the scope of large established companies and then maintain a long-term presence in these markets. First comes innovation and then we create volume.

Microscopy with high spatial and temporal resolution is a science in itself. How do you address the issue of usability for un- or not well-trained staff?
Antolovic: We offer an intuitive software interface that allows users to select the level of complexity. They can choose the advanced mode and manually finetune all parameters related to these timing fingerprints. These are not intuitive to many users. We have two modes to extract the timing information. But there is also a more encapsulated simpler interface that offers application parameters. In the specific case of fluorescence lifetime imaging, the users only need to adjust the expected range of the sample and then get an image, and that's it. To be able to use the full advantage of their camera, a certain amount of training is necessary. Therefore, we offer demo sessions.

AI has an enormous impact on the field of microscopy. Is this trend influencing your approach as a hardware developing company?
Antolovic: In most cases, we are not the final provider of a turnkey solution. Others integrate our sensors and cameras into their systems. That is why the AI-trend does not directly affect us yet. We certainly use AI to facilitate our R&D. Globally there is research on combining SPAD outputs with neural networks. Some of our colleges did so before joining PI Imaging. But I think there is still a long way to go before we really embed AI into the back of the image sensor.

Looking beyond your company, what are the most important trends in the field of 3D imaging?
Antolovic: In general, SPAD has been used for 3D imaging. Actually, this was the first application of this technology. This topic became quite popular in the middle of the last decade, but the field of 3D sensing has since then cooled down and consolidated. LiDAR, the key technique in 3D ranging, is now represented by a small number of established players providing their detector solutions to cars for autonomous driving. Yet, many car manufacturer development centers still aren’t fully convinced that this is a key technology to enable autonomous driving. I personally oppose this assessment. Perhaps we’ll see a second wave. We as a company are not focusing on 3D imaging. Instead, we see two other use cases where SPADs might soon gain momentum. The first use case is setups where power illumination is high. Due to the low light capability of SPAD technology, it is feasible to save illumination power dramatically, and also to prolong the lifetime of device batteries or to reduce the cost of very expensive illumination setups, for example, in crash-tests. And the second use case where we feel there is going to be a major impact related to high speed is movement compensation. We know photography in dark scenes has an issue with sharpness and image quality. SPADs offer high speed to oversample in time and move the images with respect to the movement and finally create high quality images in very dark scenes. This is interesting for many applications both in consumer products and industrial environments.