3D ToF SPAD Imaging - Interview with SensL

Register here for your ticket to IS Auto Europe 2018 >>


Your presentation explores 3D ToF SPAD imaging – performance and systems designs – can you give us a brief insight into what this will explore?

The current challenge for automotive LiDAR is to be able to achieve long-distance ranging of low reflective targets in bright sunlight or other harsh environmental conditions. Once these performance challenges have been met, the technology then needs to be made compact, fully solid state, and low cost. Fulfilling all of these requirements in a single system is challenging and fuelling a lot of interesting developments and innovation. 

From SensL’s point of view, the way to fulfil all of these requirements is through the use of SPAD-based sensor technology. SPADs, and the related Silicon Photomultiplier (SiPM) sensors, are intrinsically sensitive to low light levels, even down to the single photon level. This is a key property when considering the very low return signals from low-reflective targets at long distance. The next advantage is that when you have sensitivity to single photons, you can use a time correlated single photon counting (TCSPC) techniques, to enable what we refer to as the multi-shot approach to LiDAR. We believe it is the combination of single-photon sensitive sensors and the multishot approach that will really be the enabler for fulfilling the performance challenges of automotive LiDAR. 


Why do you think this is a game changer for the industry?

Beyond the performance of the LiDAR, the other challenges for a practical LiDAR system are the cost, and the requirement for the system to be fully solid state. In the context of automotive applications, solid state means ‘no moving parts’, and there are several approaches including mems and optical phase arrays. The reason this is important is that an imaging LiDAR system requires some sort of scanning in order to capture data across the field of view. Due to eye-safe laser power restrictions, flash LiDAR has limited range and many approaches to long-range LiDAR employ scanning the light from the laser and to the sensor. Although the scanning problem is still challenging, it can be simplified by only scanning the outgoing laser light and having a ‘staring’ sensor array. The columns of the array are enabled in turn to receive the returned laser light from different laser scanning locations. CMOS image sensors for these kinds of applications are already becoming a hot topic for LiDAR, but due to the advantages of single photon sensitivity, SPAD arrays are likely to be the real game changer for the industry. 


What can the attendees look to gain by hearing your presentation at the conference?

SensL will be talking about our current and future SPAD array developments for automotive LiDAR, in particular, our first SPAD array product. We will talk about the benefits of single photon detection, the multi-shot approach and the use of a scanning laser/staring sensor array configuration. We’ll also be talking about our modelling work that we use to simulate LiDAR systems of various configurations, and how we have validated this model using LiDAR demonstrator systems that we have built. Our most recent demonstrator, the Gen3 system, is a scanning LiDAR which has been used to validate the model as well as get some great images from. We’ll certainly be sharing some of the results from that during our presentation. 


Where do you foresee the LiDAR market for the automotive industry is headed in the next 5 years and why?

Fully achieving the requirements for long-range automotive LiDAR is proving challenging but progress is being made. LiDAR systems are now available for the automotive market but they are costly and don’t fulfil the requirements for fully autonomous driving. Therefore, as well as improvements in the technology to greater range and lower reflectivity targets, the cost will come down making it a more standard part that will eventually be a standard feature in all cars. I don’t think that will be the case in 5 years, but that is the way things are heading. 


What aspect of the conference are you most looking forward?

The field of automotive LiDAR is now so active and fast-moving that we are excited to see what everything is working on and the progress that is being made. Part of that excitement is being able to go along and share the progress that we have made with the new SPAD arrays and our Gen3 demo systems. Events like IS 2018 are invaluable for us to learn the from the rest of the industry, and also share our experiences in sensor technology for long-range LiDAR.