Here’s a quick look at how today’s wiring harness looks like.
Why does a wiring harness look this way? A part of that answer lies in understanding how today’s architecture has evolved over the last 50 years. As the electrical systems became more distributed, wiring harnesses also evolved to keep up.
This intricate network of wires, connectors, terminals, grommets, junction boxes, relay boxes, protectors are designed and assembled to carry amperes of current and gigabytes of data. However, they do come with some hiccups time to time.
Corrosion: This is a real problem, and probably why there are multiple DIY how-to guides on how to clean and manage corrosion in a wiring harness and articles explaining why WD-40 is your best friend. An integral part of a wiring harness are electrical contacts made of metal. And long exposure to moisture and other substances in the environment can lead to corrosion.
Corrosion has two unwanted effects. It makes the contact mechanically weaker. Secondly, it increases the resistance of the electrical contact. And increased resistance can change the electrical characteristics, and could lead to faulty sensor output or just messed up data transfer within the vehicle. Recall of vehicles due to corrosion are not uncommon see this or this.
Joints: Most of the modern wiring harnesses are point to point, which means that there are no midway wiring joints. Midway wiring joints are not desirable because it creates a point of failure. Nowadays, manufacturers use more custom designed junction boxes which makes it a lot easier to manage all the joints and splits in the harness. And junction boxes really on more durable forms of contact like fastening and sometimes even welded contacts.
Weight: A typical wiring harness weighs about 30-50kg spanning kilometers when stretched out. The bulk of it comes from the metal conductors itself, followed by insulating cover and mounting solutions.
Manufacturability: Even today, manufacturing the harness for one vehicle take at least 10 different physical handovers. Simply because of the custom nature of a wiring harness and the differences in the harness with each different model and sub-variant of the vehicle. There are processes within this that have already been automated, like cutting, creating twisted pairs etc. But some processes are still semi-automated, which means you still need an operator for the machine. End-to-end automation for manufacturing wiring harnesses is quite far away with the current state of design. There is definitely potential here from designing from scratch, given the kind of technology we have now available to us.
What’s the whole point of wiring?
But the goal of a wiring harness(the physical part of the electrical system) boil down to 2 main aspects
- Transferring electrical power
- Data Connectivity
Let’s do a quick thought exercise. If we could start from scratch, what would be efficient ways in which electrical power spread within the vehicle? Well, that is a trick question. The best actual way to transfer power is to not have to transfer at all.
Phantom feed, low energy, built-in powering are the next best options we have. A trend that is common now is having more functionality abstracted within SW. By having more powerful ECU’s, the overall number of ECU’s are dropping, as more and more functionality can be handled within the same ECU. The E/E architectures are centralizing to a small number of domain controllers and have fewer, low powered units within a small part of the vehicle, each governed by its own domain controller.
The next best thing could be if the sources of energy were distributed. And that would mean, different ECUs would have a battery within each piece of hardware. But that won’t solve any problem, as we would need to have a means of charging and maintaining them. Another key aspect is that OEMs assemble an array of ECUs from a range of suppliers.
But these are not applicable for all use cases. Here are a few factors to account for:
- Current requirement correlates to gauge size.
- Material and Length correlates to voltage drop.
- Transients, EMC, Antenna’s
- Fusing concept and the distribution of consumers.
- Single power source or distributed?
- High voltage systems to accommodate electric traction systems
The bulk of communication in today’s cars are done via CAN. A vehicle bus standard developed in the late 1980’s that have been adopted as an industry norm. But CAN has issues of limited topology options, bandwidth etc. And now we are shifting to new solutions like ethernet. But how do you choose what the next industry baseline should be?
Here’s an idea: If you have a high enough bandwidth, you can send all kinds of data on the same physical layer. Sounds familiar? Well that’s because it’s what's happening in the consumer tech world with the USB-C revolution. In addition to that, recently the EU has ruled that all future devices shall have USB-C as the main charging interface. The hard part, similar to smartphone industry, is to get a standard adopted by OEMs, Tier 1, Tier 2 etc. within a common time frame. The standards and technologies are moving too fast, that neither OEMs or even consumers for that matter want to make any long term guesses.
Not only the Physical Layer needs to be harmonized, but the connector system as well. Otherwise, we still end up in the whole USB connector fiasco that we are familiar in the smartphone world.
H-MTD, which stands for High-Speed Modular Twisted-Pair Data is an emerging connector system which is capable of transferring up to 56Gbit/sec. It has proven to have desirable EMI properties and connector manufacturers like Rosenberger have developed connectors for this PHY also with optimum mounting forces.
Another fundamental optimization is if we could efficiently transfer power and data on the same PHY. Phantom power is more common within the audio world. There exists several IEEE standards specify transferring power along with data on a twisted part cable, collectively called Power over Ethernet.
A big part of the future is more and more wireless solutions within the vehicle. The tech stack have features now that are reliable, short range, built-in security features. However it comes with a different set of failure modes. Issues like interference could become a thing. Ex: Wireless opportunities, NFC, BT, WiFi. Harmonization among these standards and frequency bands is still a necessity for a more wide spread adoption within the industry.