Autonomous racing cars have become the latest development focus for car manufacturers. University students around the world are vying for the pole position in the race for their future careers in the automotive industry by participating in international Formula Student competitions. The Aristurtle (Aristotle University Racing Team Electric) team from the Aristotle University of Thessaloniki, Greece, used the Cincoze DS-1202 to give their driverless electric race car a boost in the Formula Student competitions.

Aristurtle consists of students from different departments, from Electrical and Mechanical Engineering to Physics and Economics. The racing team already has 9 years of experience in the field of design and construction of electric racing vehicles and has received significant distinctions in international Formula Student competitions.

Cincoze provided the racing team with a robust computer vision platform DS-1202. The DS-1202 is the heart of the driverless race car and is responsible for the critical job of processing various data and sensor inputs into precise racing actions.

Computer Vision

Cameras and LiDAR are the eyes of driverless cars. They provide a clear view of the car's surroundings and help the car to recognize and detect the location, speed, and 3D shape of nearby objects. An embedded computer with NVIDIA® GeForce® GTX 1650 GPU was needed to process data from the sensors and power the neural network inference and model predictive control optimization.

Abundance of I/O

The student team looked for computing hardware to connect the following peripherals to the APU (Autonomous Processing Unit): multiple USB ports for cameras, legacy COM ports for serial communication, two Ethernet ports for LiDAR and telemetry, and two Mini-PCIe for the APU’s CAN controller board and wireless connectivity. DIO was also required for connecting external LEDs for important ROS signals, such as velocity estimation and hardware status.

Thermal and Anti-shock

The students considered acquiring a computer with robust design. With the intensive CPU and GPU computing, heat dissipation becomes very important to avoid processor throttling. Additionally, rugged industrial features like anti-shock and -vibration design are required to withstand the severe vibrations during racing.

Reliable GPU Computing

Cincoze DS-1200 series is powered by a 9th/8th Gen Intel® Core™, Pentium®, or Celeron® CPU. Combined with an NVIDIA® GPU in the PCIe expansion slot (up to 75 W), the DS-1202 is perfect for AI applications such as autonomous driving, visual inspection, surveillance, and deep learning. As for the thermals, an external fan on top of DS-1202 creates good airflow and good cooling.

Rich I/O and Modular Expansion

The DS-1200 comes with rich I/O interfaces, including 2x GbE, 8x USB, 1x DVI-I, 2x DisplayPort, 2x RS-232/422/485, and a remote power/reset connector. Through Cincoze’s innovative CMI & CFM modular technology, users can add extra I/O and functionality, such as GbE/PoE, serial ports, optical isolated digital I/O, and power ignition sensing function. In addition, the DS-1200 series provides 3x full-size Mini PCI Express slots for wireless connectivity or CAN-BUS expansion for vehicle communication. There are also DS-1201 and DS-1202 models with one and two PCIe expansion slots respectively.

Rugged Design

The DS-1200 is built with a rugged extruded aluminum chassis. It features fanless, jumper-less, and cable-less design while providing a wide operating temperature range of -40°C to 70°C, a wide power input of 9V to 48V, and high tolerance of vibration/shock (5G/50G). It has passed industry compliance EN50155 (EN50121-3-2) and IEC/EN62368-1 certifications for operations in harsh environments.

In addition to the DS-1200, Cincoze recently launched the DS-1300, a powerful and expandable embedded computer powered by a 10th Gen Intel® Xeon® or Core™ CPU and support for one or two PCI/PCIe slots for applications requiring high computing power in harsh environments.