Page 20 - EE Times Europe Magazine – June 2024

20 EE|Times EUROPE

AUTONOMOUS VEHICLES | SENSORS
New Developments in MIPI’s High-Speed

Automotive Sensor Connectivity Framework

By Ariel Lasry, MIPI Alliance

Higher link speeds, security and a compliance program will connect to all the vehicle image sensors
have been added to the MIPI Automotive SerDes via a small number of zonal ECUs, not only
enabling software-defined vehicles but also
Solutions framework. allowing wiring harnesses to be simplified for
weight and cost savings.
Automotive image sensors, which include cameras, To support these evolving E/E architec-
LiDAR and radar, play a critical role in advanced tures, MIPI developed MIPI Automotive
driver-assistance systems (ADAS) and autonomous driv- SerDes Solutions (MASS), a framework for
connecting sensors and displays over long
ing. Today, the systems that deliver advanced SAE Level 2 cable lengths with functional safety (FuSa)
features typically utilize a combination of up to 10 cam- and security built in. Since its initial release
era and radar sensors. This number is expected to go up in 2020, several components within the
significantly to enable higher levels of automation, poten- MASS framework have evolved, and in 2024,
tially requiring as many as 45 image and other sensors to major advancements include the addition of
deliver the most advanced SAE Level 3 (and beyond) features. As the number of image sensor security functionality, an even
faster MIPI A-PHY v2.0
image sensors in vehicles has increased, the MIPI Camera Serial Interface 2 serializer/deserializer (SerDes) physical
(CSI-2) standardized image sensor protocol has become the core imaging proto- layer and the further development of an
col used within automotive systems. A-PHY compliance program.
The need to integrate these greater numbers electrical and electronic (E/E) architectures COMPONENTS OF THE MASS
of image sensors, with ever-higher resolu- that leverage the latest high-speed automo- FRAMEWORK
tions, creates multiple design challenges, tive networking protocols. Representing a collaborative cross-industry
including connecting multi-gigabit image To enable high-speed, software-oriented effort to meet the needs of the automotive
data streams to application processors and architectures, the automotive industry is industry, MASS consists of four principal
processing the huge volume of raw image data evolving away from the domain-specific components that, when combined, create an
generated by the sensors. To meet these chal- architectures of today and toward cross- end-to-end high-speed connectivity frame-
lenges, vehicles are becoming increasingly domain centralized architectures. These cen- work for sensors and displays:
software-oriented, enabling sensor fusion, tralized architectures use just a few powerful • The reuse of widely adopted
where powerful algorithms combine data computers compared with the high number higher-layer protocols: MASS leverages IMAGE: ADOBE STOCK
from multiple image sensors into a single data of individual electronic control units (ECUs) a suite of higher-layer application proto-
model of the environment around a vehicle. used in today’s domain-specific architec- cols used in billions of devices and already
The industry is also developing innovative tures. Several powerful vehicle computers widely adopted in automotive. These

JUNE 2024 | www.eetimes.eu