Unmanned Systems Technology 018 | CES show report | ASV Global C-Cat 3 USV | Test centres | UUVs insight | Limbach L 275 EF | Lidar systems | Heliceo DroneBox | Composites

75 Lidar | Focus simplifies the component procurement and supply chain. A single-detector system can be used for applications requiring a 0-20 m range, such as level sensing, security and surveillance, and proximity detection by using various beam options. A multi-element platform (typically containing a photodetector with 16 elements) is used for applications requiring a 0-50 m range, and provides lateral discrimination of objects, simultaneous and independent acquisition capabilities in each element and various beam options. Solid-state systems There is also a push to combine all the elements of a laser sensor in a single package. While Flash arrays can be considered solid state as there are no moving parts, another solid-state approach combines a 900 nm laser source that feeds a series of steerable channels in silicon. Varying the voltage across the channel changes the refractive index of the silicon channels and so varies the phase of the light to steer a beam in the same way that phased array radar systems use steerable radio beams. A single laser source is used to avoid phase variation and maintain coherence of the laser signal across the channels. This allows the beam to use a small spot to zoom in on a particular area, typically 3.5 cm across at 100 m, which means it can see the shape of a body and movement of a hand. The spot can be larger, typically 60 cm at 100 m, for faster scanning of wide areas, and the range is typically up to 150 m, depending on the balance of the laser’s output power and the sensitivity of the receiver array. Receivers are built into the silicon, but instead of the eight to 32 receiver elements in a rotating scanner there are a million in a 1000 x 1000 array, similar to the 3D Flash sensor approach. The sensing array operates in the Geiger mode that detects individual photons. Each sample is sent as a coded (8- to 11-bit) pulse train that helps identify the direction for the return path, and can detect as little as a single photon. Using a coded pulse train also avoids any problems with interference from other light sources, and multiple trains can be used for longer range. Using large, 300 mm silicon wafers with a relatively mature, less expensive 35 nm process keeps costs down. By comparison, volume chip manufacturing is at 28 nm and leading-edge manufacturing is at 10 nm. Micro-mirrors are made on a much coarser, micron- level process. The solid-state approach allows thermal compensation of the components, which is a key issue with all the laser systems and allows a solid-state device to ship without active thermal control. However, all this requires high- precision assembly equipment to mount the gallium arsenide or gallium nitride laser diode on the channel steering substrate. The aim is to load a cassette of 300 mm silicon wafers into the front of the manufacturing process and get a fully assembled Lidar sensor at the end. Because the steered beam provides a scan, the output of such sensors is in the same standard Lidar format as existing sensors. Mirrors still performing Despite advances in other areas of Lidar, scanning mirror systems are still being used and developed, but the focus has changed to more static, high-resolution applications such as traffic monitoring or robotic systems. This can generate larger, higher resolution 3D point clouds with multiple layers of information but requires higher data rate outputs such as gigabit Ethernet. This data can then be processed in the cloud and sent to individual vehicles. Lidar scanners have been proposed for traffic lights to provide data in this way. However, most developments for onboard mobile applications are moving to the more integrated solid-state or micro-mirror array architectures. Unmanned Systems Technology | February/March 2018 The first samples of the LeddarCore LCA2 System-on-Chip, a highly integrated Lidar engine, and its implementations into a 3D Flash Lidar module targeting mid-range applications were shown in January 2018. The LCA2 is scheduled for volume production in 2019 and mass commercial deployments by automotive manufacturers in 2020. The LeddarCore LCA3, currently under development with the first samples available in late 2018, is targeting higher levels of automation and will provide an effective range of up to 300 m (Courtesy of LeddarTech)