How Tesla is Replacing the Age-Old CAN Bus

Every Tesla vehicle is filled to the brim with modern and advanced features - and there is a massively complex network of devices powering that - from the FSD and infotainment computers, to the various networked sensors throughout the vehicle.
That massive network of wiring is traditionally run on a system called CAN, or the Controller Area Network - which was developed by Bosch all the way back in the 1980s. Since then, it has been the industry standard for in-vehicle part-to-part communication for decades.
However, just like the horse and buggy, it may be time for CAN to be put out to pasture as it struggles in the data-driven modern environment. Massive amounts of sensor data, high-resolution infotainment screens, over-the-air (OTA) updates, and centralized Electronic Control Units (ECUs) mean that the old standard just can’t keep up anymore.
Tesla is now actively developing and deploying a next-generation vehicle network to replace CANBUS, and this new network will likely function in synergy with the move to the new 48-volt low-voltage architecture being pioneered by the Cybertruck.
CANBUS - The Old Workhorse
CANBUS was originally developed in 1983, released in 1986, and then standardized by the International Standards Organization (ISO) as ISO 11898 in 1993.
It’s a venerable standard that was revolutionary at the time, as it drastically reduced wiring complexity compared to the point-to-point methods being used in the late 80s and early 90s, and saw immediate mass adoption across the entire industry.
CAN is a message-based protocol, where nodes broadcast data with identifiers. The priority of packets determines their movement and access. However, CAN 2.0 and CAN FD are both extremely limited - CAN 2.0 is limited to a glacial 1Mbps, and ~8Mbps for the more “modern” CAN FD.
CAN FD barely makes the mark for 1080p video streaming at 60fps - if it is pre-encoded. Unencoded raw video surpasses what CAN FD is capable of, and greatly limits its capabilities and usages in a modern data-first vehicle like a Tesla.
CAN is also complex - it is simpler than a point-to-point wiring system, but the multiple CAN buses and gateways result in a complex, heavy, and costly wiring harness that can be next to impossible to diagnose, repair, or replace.
Tesla’s Next-Gen Networking
Tesla’s next-gen networking is all about timing - and unlike CAN, where two messages coming in at the same time can collide (resulting in neither reaching the node), Tesla’s TDMA, or Time Division Multiple Access, assigns specific time slots. This means that access to each node or data point is guaranteed and avoids interference.
You can think of CAN being like everyone yelling in the same room - but TDMA being a tightly scheduled series of one-on-one meetings.
However, TDMA isn’t just a simple sorting system. According to Tesla's patent application, the network operates in repeating cycles. At the start of each cycle, a Network Allocation Map (MAP) is transmitted. Think of this MAP as the dynamic schedule for that cycle – it tells every node exactly which time slots are reserved for which communications. Each reservation specifies the transmitting node, the receiving node, the duration of the slot, and, crucially, the type of traffic it is for.
This allows for sophisticated Quality of Service (QoS) management, separating data into different categories. The patent specifically calls out two main types:
Low Latency (LL) Traffic: These are for critical, time-sensitive signals (think sensor readings for FSD, airbag triggers, control commands). They get assigned short time slots that repeat very frequently within the TDMA cycle (potentially every 500 microseconds, according to one example in the patent) to guarantee delivery within a strict maximum delay. The data packets themselves are kept small, maybe only tens of bytes, to fit these quick slots.
Bulk Traffic: This is for data where total volume is more important than millisecond-level delay (think infotainment data, camera video feeds, maybe larger data logs). These get assigned longer time slots, allowing for larger data packets (over 100 bytes in one example), ensuring high overall throughput even if they don't repeat as often as the LL slots.
This whole system relies on precise synchronization across all nodes. The patent mentions synchronization signals within the TDMA cycle and specialized modem hardware to keep everything perfectly timed.
The network can also be structured into logical domains (like front-left, cabin-right, etc.), each managed by a Domain Master node that handles the MAP and communication within that zone. So, TDMA isn’t just a sorting system; it's a highly managed network implementing traffic prioritization (LL vs. Bulk), dynamic slot allocation via the MAP, and potentially managed by centralized Domain Masters, all designed for efficiency and reliability.
48-Volt and LVCS
Many of these networking concepts appear designed to work hand-in-hand with Tesla’s recently-released LVCS - or Low Voltage Connector Standard. LVCS simplifies vehicle wiring networks by drastically reducing the number of connector types needed from over 200 down to just six. While the patent focuses on the data protocol, LVCS simplifies the physical layer, and the 48V architecture it's built on also enables using the vehicle's DC power lines as a potential network medium (PLC), helping to reduce complexity.
Tesla has been utilizing these new approaches in the Cybertruck, as evident in their new and unique interactive wiring diagram, which helps technicians debug wiring issues. We can expect even more features to take advantage of the new capabilities in the future.
48V also means thinner wires, which reduces costs, and LVCS simplifies the connectors on both the harness and nodes - which means less part complexity, further simplifying the manufacturing and supply chain, while also ensuring vehicles are more repairable.
Wrapping Up
This is another innovation that Tesla is introducing to its fleet - and while we initially looked at it and thought, “Wires? How boring,” we soon realized that it is, in fact, the skeleton that Tesla will use to build its future systems.
That means smoother, faster, and more robust FSD data transfer within the vehicle, resulting in snappier and more effective decision-making. A quicker and more functional infotainment system and better support for deep-reaching OTA updates due to the reduced internal complexity and lack of reliance on internal CAN buses, which couldn’t be updated.
This is a massive technological leap over the decades-old CAN bus system, and while it may be invisible to the average user, it is an excellent example of all the engineering that goes on in under the hood of every Tesla vehicle.