Unmanned Systems Technology 025 | iXblue DriX I Maintenance I UGVs I IDEX 2019 I Planck Aero Shearwater I Sky Power hybrid system I Delph Dynamics RH4 I GCSs I StreetDrone Twizy I Oceanology Americas 2019

90 are packaged in a Bosch BMC050 unit, and an STMicroelectronics L3GD20 is used as the gyroscope. “As the unit gives you magnetic strength in three different directions, you can even use it as a compass if you wanted to, which is something we’ve done for some little side projects,” O’Sullivan adds. Software The StreetDrone Twizy’s most popular self-driving software application is the Autoware Foundation’s open source algorithm stack, Autoware.IO, as StreetDrone is a founding member of the group and advocates its accessibility to its customers. It also means StreetDrone’s engineers are continuing to contribute developments and improvements to this software alongside those from its customers. As O’Sullivan explains, “Within that, we have created our vehicle in a simulation environment, so customers can perform dynamic simulations using it with all the vehicle dynamics already calculated for them. That integrates nicely with Autoware’s ROS-based software as well.” As a function of the company’s desire for ease of use, and its anticipation of growing autonomous mobility-on- demand services, the StreetDrone Twizy is compatible with mission control and fleet management software from Swiss start-up Bestmile to track conditions and designate the locations of connected vehicles. “We’ve worked with their engineers to integrate Bestmile’s cloud-based services into our system, so our end-users can monitor their vehicles right off the bat,” O’Sullivan says. “It can also respond to taxi requests by judging which vehicle is best placed to reach a given location, and coordinate all the vehicles collectively, rather than having a bunch of vehicles acting independently of one another.” The self-driving and fleet management software stacks are integrated with the drive-by-wire control algorithms and safety systems into XenOS. This is a single operating system software package to ‘simplicate’ the end-user’s interface with their StreetDrone vehicles. “The way self-driving car development works at the moment is that if you have 50 different mobility services you want to integrate into your business, you have to make arrangements with 50 different companies and sign 50 different contracts, just to get your car to the level of capability you want,” O’Sullivan says. “Our approach though would mean you’d need to link only once with XenOS, then that operating system fills itself with the rest of the services and technologies based on the contracts we’ve gone through ourselves.” The road ahead StreetDrone is also developing several new vehicle models for specific end- users, each of them with more range and passenger space than the Twizy, and all with electric powertrains. They are likely to be announced later this year. “We’re also open to working on hybrid and non-electric vehicles as well,” O’Sullivan says. “As I said, our software and competencies are flexible, and we’ve just not had demands for anything but electric platforms.” Between lines of code and production car models, StreetDrone’s niche is unquestionably fulfilling a significant need among self-driving automotive researchers, to the benefit of tomorrow’s roads and highways. April/May 2019 | Unmanned Systems Technology Digest | StreetDrone Twizy Drivetrain: fully electric Seating: two Endurance: 64 km Maximum speed: 80 kph Charge time: four hours (120 V AC) Some key suppliers Lidar: Velodyne Lidar: Robosense Cameras: Point Grey Cameras: Basler Radar: SmartMicro Radar: Continental GNSS-IMU: PEAK-System Processors: Nvidia Base self-driving software: Autoware Fleet management software: Bestmile Operating system: in-house Sensor and computer mounts: in-house Chassis: Renault Powertrain and power electronics: Renault Specifications Bestmile’s mission and fleet management software enables tracking and control of multiple Twizy vehicles across cities, for groups such as autonomous taxi developers seeking to test and optimise collective navigation algorithms