Advanced Analysis in Heavy Logistics Road Transport

I want to consider Advanced Analysis (AA) in heavy transport. Sounds complicated, doesn’t it? Advanced Analysis (AA) is the use of computer intelligence on the ground to improve safety, productivity, delivery timeliness and reliability, by using computers to predict the best decisions in real time.

A modern heavy vehicle (including its trailers) has many on-board computers. There are more computers back at base. They monitor data and try to predict outcomes or control them. It is very likely that the complexity of the systems and the volume of the data they produce will overwhelm the human operator. The development of increasingly smarter computer intelligence at base offers simplification and understanding from all this data. None of this comes free because Advanced Analysis relies upon rules and the rules must be developed by humans. But I expect that machine learning will improve the rules and tailor them to particular fleet operations.

Let’s identify the intelligent subsystems that a modern truck probably has. The on-board subsystems that are controlled by a computer are shown in the figure. They probably talk to other computers on a CAN network so that some operational data of the vehicle is shared. Each of the sub-systems performs a control function based upon local sensor data and shared operational data. So, for example, the wheel speeds are calculated by the ‘Electronically Controlled Brake System’ which is ‘local data’. The ‘Vehicle Stability Control’, which oversights the brake system needs to know the engine speed, and it must be able to control it. The camera system that may be installed to improve driver visibility around the truck, can give alerts to the driver, but probably does not need to talk to the engine. Therefore, some operational data is shared, and some is ‘local’. A modern truck has a CAN data network with several hierarchies in the CAN network. The Figure shows some of the systems that could be installed; but not all.

The Robert Bosch company developed the Controlled Area Network (CAN) and patented it in 1983. It was a relatively low-speed computer communication network intended for use on vehicles, so that additional computer control modules could be added without any programming or reconfiguration being needed. This was possible because the modules were manufactured with the CAN message format set and with a method for resolving clashes built into the protocols. Effectively, each module knows what to say, how to talk and how to grab the microphone! Each module will also listen to the operational data that was being broadcast. The message structure is described in SAE standard J1939 (29 bit message identifier and up to 8 data bytes – CAN 2.0B frame format) and ISO standard physical layer specified in 11898:2023. The communication bit rates are 250 kbits/s and 500 kB/s.

I won’t go any deeper into how CAN works and will finish with the following summary:

All modern heavy trucks have a CAN communication network. The basic channel has a two-wire twisted pair that is terminated at each end by 120 ohm resistors. Computer modules containing CAN transceivers are ‘teed’ into the data bus. This is called the physical layer. It can be 40m long and contain 30 computer modules, called nodes.
CAN allows computer modules communicate operational data and messages in a format described in standard SAE J1939.
There can be a mixture of low and high-speed CAN data-buses on a vehicle. The powertrain CAN data-bus is usually high speed whereas the instrumentation CAN might be slow.
There is usually a ‘gateway’ CAN module that interconnects the multiple CAN data-bus on a vehicle. This is often the body controller in the cabin. It provides a level of isolation between different CAN data-buses and directs appropriate data between them.
The CAN on North American vehicles is referenced to 12V whereas on a European truck it is referenced to 24V. This should not matter because the physical layer standards around the world use the same communication voltages (High: 3.5V Low: 1.5V).

Figure: Some of the data flows in from a modern heavy truck. Image: ARTSA-i.

The message types and structure have been harmonized between USA and international standards via the SAE J1939 standard. However, manufacturers do use different length message structures, and it is unlikely that a node made in Europe will seamlessly work on a North American truck, and vice-versa.
Non-original CAN nodes are not usually connected to the OEM manufacturers’ CAN network, but they could be. The gateway module may not pass the information on, but the module could obtain operational data. OEM manufacturer’s would probably not want non-original nodes added to a CAN data bus.
External modules such as telematics units are usually connected via the OBD II diagnostic plug. This is in the cabin (usually close to the steering column). The connection is via a standard connector type.
The brake CAN on a European and Japanese truck is available on pins 6 & 7 of the trailer EBS connector (ISO 7632).
North American trucks do not provide the brake CAN to the trailer, because there is no brake CAN. That is, North American trucks do not have brake module CAN communication.
Telematic systems do not communicate to the outside world via CAN. They use a telecommunications network.

The transmission of vehicle data between the truck telematics system and home base is likely to be via satellite or mobile phone tower communication. Integration of this data into the operator’s business system is now routinely done. Systems that are used to justify compliance with regulations should meet standards developed by the Transport Certification Agency (TGA). TGA is the owner of the National Telematics Framework that underpins the use of operational vehicle data to meet regulatory, contractual or commercial requirements. This Framework has been incorporated into ISO standard 15638, which signifies the TGA’s world leadership in this domain. This work has assisted Australia to be world-leader in the interconnection of vehicles to business systems and regulators accreditation schemes.

The Figure shows the possibility of Vehicle-Infrastructure communication, which has not been implemented here (or in North America). It is probably a 20-year project that is yet to start. The Figure also shows OEM (static) access to module data via the Internet of Things (IoT). This will raise issues of data ownership when an incident occurs.

The full promise of Advanced Analysis is yet to be realised. The decisions that could come from analysis of real time vehicle data off the vehicle are: