While the for fully terms of implementation Autonomous vehicles is constantly changing, this doesn't slow down the growth of the automotive startup ecosystem. That some companies are still working over achieving the goal of level 5 autonomy, while others are addressing issues related to electric vehicles, cybersecurity, processor equipment, and other painful moments in the automotive industry.
We invite you to get acquainted with the newest automotive startups of 2019 who are already impacting the automotive industry.
AEye: Many self-driving and networked vehicles rely on lidars for navigation. But lidars alone are not enough. Aeye's solution, iDAR (short for intelligent range detection and control), addresses the processing time and processing power issues inherent in lidars. iDAR combines solid-state and flexible lidars, and it also has a built-in HD camera that works in low-light conditions. The logic of iDAR operation is provided by AI-based computer vision, which accelerates the operation of the vehicle's perception system. According to the company, iDAR can increase the speed of perception up to 10 times and reduce power consumption by five to ten times. In recent performance tests tracked and verified by VSI Labs, one of the country's leading automotive technology specialists, AEye's iDAR system detected and tracked a truck at a distance of one kilometer, which is four to five times the distance that modern lidar-based systems operate at. And more info...
Affectiva: Affectiva's goal is to add an emotional component to your car. The company has developed an AI system that can recognize a range of human emotions, including anger and sleepiness. The AI works using a system of video cameras installed in the car interior. The ultimate goal is to allow cars to respond in a variety of ways based on the emotions of the driver and passengers. The car can automatically adjust the temperature or play soothing music depending on the driver's mood. And in extreme situations, a driverless car can pull over and alert an ambulance if it detects that something has happened to the driver-such as a heart attack or loss of consciousness.
Byton: One of the favorites of the public at the 2018 LA Show, Byton, does not refer to the automotive business, but to the business of creating "mobile platforms". The company is developing self — driving cars that are less conventional vehicles and more mobile offices and workplaces. In addition to the eye-catching aesthetic that captivated visitors to the LA Auto Show, one of the main features of their M-Byte concept car is the all-glass cockpit, which replaces the buttons and knobs on the dashboard with a fully on-screen interface.
The glass cockpit allows you to control all the functions of the dashboard and much more with gestures, voice, and even breathing. The LA Auto Show also featured a second Byton concept car, the K-Byte, a level 4 driverless car. The Chinese company is currently testing M-Byte in real-world road conditions and has opened offices in Los Angeles, Santa Clara, Munich, Hong Kong and Beijing with the aim of introducing its vehicles to the US and Europe over the next two to three years.
Carmera: Carmera focuses on the infrastructure needed to help driverless vehicles navigate space as accurately as possible. The company has developed a set of HD maps for Autonomous cars. Carmera uses crowdsourced data from vehicle sensors to create a constantly updated 3D map for driverless cars that can be used for navigation, analysis, infrastructure planning, and fleet tracking. Carmera was among the top three finalists in the "Top 10 automotive startups" competition at the 2018 Los Angeles auto show. Earlier this year, Carmera announced a partnership with Voyage, in which it will provide mapping services for driverless taxis in Florida.
DarwinAl: You can't talk about self-driving cars without discussing artificial intelligence, namely deep learning. All these neural networks for vision control, natural language processing, and other functions required for self-driving machines require a lot of computing power, as well as a large amount of memory. DarwinAI is a deep learning software company that wants to make deep learning less resource — intensive by optimizing algorithms and neural networks.
In short, feed DarwinAI a neural network, and it will return you a simpler, but equally effective version of it. The proprietary optimization method, the Generative Synthesis platform, goes beyond simple compression and uses methods to reduce weight and reduce its accuracy to create more compact neural networks. This approach also increases what the company calls the "explainability" of the neural network, making it easier for developers to understand how the neural network functions and why it makes certain decisions. Because the company's methods can be applied in various industries, DarwinAI's work is interesting for the driverless car market, where efficiency, accuracy, and speed are critical to the software systems that underpin the car.
Electric Vision Aircrafts (EVA): Maybe we don't need roads where we're going? The X01 from French startup Electric Vision Aircraft (EVA) isn't a car or truck, it's a small electric plane — essentially a flying machine. Capable of vertical takeoff and landing, the X01 is a two — seat electric aircraft that can reach speeds of 400 kilometers per hour (about 249 miles per hour). It is also equipped with radar, lidar, and 12 cameras for fully unmanned flight (the maximum speed in unmanned mode is 250 kilometers per hour or about 155 mph). According to the company's specifications, the X01 is compact enough to fit in any Parking space (it has dimensions of 5 x 2 meters with folded wings), and works silently, like an electric car – during the flight it is absolutely not heard. Safety features also include a built-in parachute. The X01 is not yet in production and no official release date has been announced, but the company is currently accepting pre-orders for the vehicle from any customer willing to pay approximately $284,000.