FlexRay
History
The FlexRay consortium was founded in 2000 by BMW and DaimlerChrysler, in collaboration with semiconductor manufacturers Philips and Motorola. The goal was to develop a deterministic, fault-tolerant communication protocol capable of meeting the requirements of safety-critical automotive systems (x-by-wire).
The consortium expanded rapidly: General Motors (2001), Ford (2002), then Robert Bosch, Volkswagen, and more than 120 companies. The FlexRay v2.0 specification was finalized in 2004, and the first production vehicle equipped with it was the BMW X5 (E70) in 2006, followed by the BMW 7 Series (F01) in 2008 with full use of the protocol.
The consortium was dissolved in 2009 after the specifications were transferred to ISO under reference ISO 17458.
Technical specifications
| Bit rate | 10 Mbit/s per channel (20 Mbit/s in dual-channel mode) |
| Architecture | Dual channel (A and B) for redundancy and fault tolerance |
| Access method | TDMA (Time Division Multiple Access) — deterministic, collision-free |
| Frame size | Up to 254 bytes of payload data |
| Cycle time | Typically 1 to 5 ms |
| Synchronization | Global time base, synchronous architecture |
| Topology | Bus, star, or hybrid |
Structure of the communication cycle
The FlexRay cycle is divided into four parts:
- Static segment (mandatory) — fixed-duration slots, one per ECU. Each node transmits in its assigned slot, guaranteeing deterministic behavior for safety-critical messages. This is the core of the protocol.
- Dynamic segment (optional) — an event-driven portion using FTDMA (Flexible TDMA) with mini-slots. Allows non-critical messages to be transmitted on demand, with lower priority.
- Symbol window — used for network signaling (wake-up, etc.).
- Network Idle Time (NIT) — the period used to synchronize node clocks.
This hybrid structure (static + dynamic) makes it possible to combine the predictability of TDMA for critical functions with the flexibility of event-driven communication for secondary data.
Fields of application
- X-by-wire — steer-by-wire, brake-by-wire, drive-by-wire: applications where a network failure would have safety-critical consequences.
- Chassis control — adaptive suspension, stability control, adaptive cruise control (ACC).
- Powertrain — coordination between the engine, transmission and hybrid systems.
Why FlexRay lost momentum
Despite its technical qualities, FlexRay did not achieve the widespread adoption that was expected, for several reasons:
- High cost — FlexRay transceivers and controllers were significantly more expensive than their CAN equivalents.
- Integration complexity — configuring TDMA slots and global synchronization required specialized expertise and tools.
- Competition from automotive Ethernet — Automotive Ethernet offers higher bit rates (100 Mbit/s to 10 Gbit/s), a richer software ecosystem (TCP/IP) and falling costs, while providing determinism via TSN.
- CAN FD as a compromise — for many applications, CAN FD (8 Mbit/s, 64 bytes) bridged the performance gap between classic CAN and FlexRay, at a much lower cost.
FlexRay is nonetheless still used in high-end vehicles (BMW, certain Audi and Mercedes models), and its deterministic communication principles have influenced the development of TSN for Automotive Ethernet.
