Bench Talk

There is more than a possibility that the future autonomous cars we drive could be designed by some of the students involved in this year’s NXP Cup EMEA—as they have demonstrated ample talent and ingenuity over the course of the contest. After six qualification events spread out across the whole of the EMEA region—in the Czech Republic, France, Germany, Lebanon, Morocco, and Romania—the 19 top teams (from a record 160 that originally enrolled) progressed to the final. This was once again held at the Fraunhofer Institute IIS, in Erlangen, Germany, on April 29^th and 30^th.

The teams were eager to show the latest updates they had made to their 1/18-scale autonomous cars, and tackle the four demanding track challenges set by the NXP Semiconductor project management team. The winners of the EMEA title will then get the opportunity to go to the NXP Connects event (in Santa Clara, California) in mid-June.

NXP Cup Racecar Design

I was certainly impressed with the new sensor additions made to the vehicles to help manage the obstacle avoidance and speed control challenges. Team GRKBrain (who hail from Kosovo) continued to fine-tune their algorithms created for the NXP Cup qualification heat in Munich, using their SlamTec A2 RPLIDAR interfaced with NXP’s standard-edition FRDM-KL25Z ARM^® Cortex^®-0+ microcontroller. Their performance in the obstacle avoidance section was unsurpassed. Team SlowFox from Masaryk University (in Brno, Czech Republic) opted for improving their computing power by upgrading to NXP’s FRDM-K66F. They continued with the tactic of relying on Mbed™-certified boards for the ease of programming. Sensor fusion that brought data together from a 3-axis FXAS21002 digital gyroscope plus the magnetometer and accelerometer data derived from a 6-axis FXOS8700CQ enabled better track handling. The team also used rotary encoders to ensure proper reading of the speed of the car’s two brushed motors. This translated into enhanced speed control. It also delivered improved electronic differential, which meant the vehicle was more adept at negotiating the curves of the track. Other cars, like the one developed by team Hall-9001 (from University of West Attica, in Greece), opted for using a Parallax 28041 optical sensor module to take care of the obstacle avoidance aspect, along with Infineon TLE4935L Hall sensors for speed control.

Figure 1: Members of Polish team KAW4Wheels. (Source: Mouser)

As the optional NXP Cup challenges of obstacle avoidance, the speed control zone and the figure-eight track require high-precision driving, team KAW4Wheels (from AGH University of Science and Technology, in Krakow, Poland) went with a custom-designed board. This featured an NXP MKV58F1M0VLQ24 ARM^® Cortex^®-M7 core clocking at 240MHz. Its Floating Point Unit (FPU) afforded better calculation precision and provided greater speed than the FRDM-KL25Z board that is part of the standard car kit. The Cortex^®-M7 core offers the processing power to deal with the signals received from two encoders (one being associated with each brushed motor). An array of four HC-SR04 ultrasonic sensors from OSEPP Electronics were to prove pivotal in steering past obstacles, as well as providing accurate operation for the speed control zone. The team took second place in the figure-eight challenge, and were third overall.

The big winners of the event, however, were the two teams representing Haute-Ecole Arc Ingénierie (from Switzerland) as they fought one another head-to-head for the title of 2019 EMEA champion. The cars from both teams (referred to as ArCar1 and ArCar2) had a very similar form, with each featuring a specially customized board. ArCar2 possesses a pair of Melexis MLX75306 CMOS cameras framed by four red LED emitters. These helped reduce the impact of any ambient light variation during the race. One camera was focused on the long-range vision of the track ahead (providing more details about the possible speed it could reach), while the other dealt with what was going on at close quarters (providing image data from the front of the car, to aid steering). The combination of the signals coming from these devices was then processed via an NXP Kinetis MK64FN1M0VLL12. For the speed-control sensing, a Honeywell SS443R sensor was employed. During the training sessions, a Redspine RS9110-N-11-22 Wi-Fi module enabled the retrieval of diagnostic data.

Figure 2: Matthey and Arfa from ArCar2. (Source: Mouser)

Unfortunately, Team ArCar2, which was comprised of Matthey Johan and Arfa Hakim, did not manage to score any points on the figure-eight challenge, as their car kept going off the track. However, they scored well in all the other disciplines—particularly in the speed race, where they clocked up a time of 23.1s—and this enabled them to take second place overall. They received a special prize from Mouser, made up of a multitude of electronic boards to assist them in preparing for next year’s competition.

Figure 3: ArCar2 in the speed trial. (Source: Mouser)

NXP Cup EMEA Winners

The EMEA winners were Romain Donzé and Alexandre Mäder with ArCar1. The vehicle was virtually the same as that of their counterparts—the only notable exception being that it used just two LEDs (rather than four) to illuminate the track during the race. It got around the speed trial course in a pacey 20.3s, as well as scoring maximum points on each of the supplementary challenges, to give the team the undisputed title for this year. When interviewed following the event, members from the two Swiss teams said that a key element for them had been learning how to use the features of each component on the car to their fullest, and to squeeze out everything they could to ensure the best possible performance.

Figure 4: EMEA winners Romain and Alexandre of ArCar1. (Source: Mouser)

Follow the NXP Cup Competition

You can watch the recordings of the NXP Cup EMEA finals on the Fraunhofer IIS YouTube channel, with both the speed race and the additional challenges covered. To keep posted on developments for the 2020 NXP Cup EMEA season, please follow Mouser Electronics on Twitter.