The heavy truck industry is inundated with emerging, innovative technologies in the vehicle, the trailer and mounted special equipment. It’s like a migration of a wide spectrum of technologies for almost every major component. But it’s important that we not “outrun our headlights,” because many new arenas of technology are progressing rapidly and emerging before needed standards can be developed and agreed to by the industry.
Are we evolving toward a high-speed/wideband local area network that is moving down the highway at 65 mph? To what extent do we need to be concerned about in-vehicle data communication standards as we move forward?
History says that as an industry, we do feel technology standards are very important. The existing SAE J1939 and datalink is the result of work initially started over 30 years ago. What do we need to stress so that the next 30 years of in-vehicle technical evolution will be as valuable as the last 30?
At the annual SAE Truck and Bus meeting in 1988, an important event for the heavy truck industry occurred:
Indianapolis, November 7, 1988 – A new plan for uniform truck electronic service and diagnostic systems was unveiled today by an ad hoc group of leading companies in the industry. (Ryder, Cummins, Caterpillar, Freightliner and Microflex). If adopted by the industry, the plan could save truck manufacturers, fleet operators and shippers substantial sums through the standardization of diagnostic and information systems now being built into trucks. The proposal guidelines involve diagnostic, information and display functions in (electronically controlled) components covering engine, transmission, brake, electrical, heating, cooling, lighting and dashboard display systems on board commercial vehicles.
This press conference was accompanied with a booklet entitled Recommended Practices Relative to Vehicle Diagnostics for the Heavy Truck Industry. The booklet laid out the details that ultimately resulted in SAE standards J1587 and J1708.
At the time I was CEO of Microflex, which served as the holder of the baseline technical recommendations and held the core engineering team.
This was 32 years ago. The original data links have evolved through several advancements to the current J1939. Plus, in today’s heavy truck productions, there are multiple proprietary CAN buses in the heavy truck as well.
I have spoken to several of the engineers that worked on that original specifications from 30+ years ago, as well as some that are working on next-generation data links for the heavy trucks of the future. Did we hit the target? What do we have today? What should we be doing over the next 30+ years?
Did we Hit the Target?
The general consensus is that we did hit the target which was envisioned by that original work.
At the time, we had multiple primary components on the vehicle that were electronically controlled, but there was no common communication nor an adaptable way to retrieve diagnostic information in a standard way. Both of these issues were addressed with the standards J1708/J1587.
Fast-forward 30 years and one could say that yes, the J1939 datalink does in fact provide diagnostic information for electronically controlled components on the vehicle. Even with the proprietary CAN networks embedded in J1939, just about all of the component manufacturers post diagnostic/fault codes up to J1939.
However, as a result of the proliferation of electronically controlled components, there are today literally tens of thousands of combinations of components and fault codes that can be posted.
Guy Rini, who formerly headed up advanced engineering for Mack and Volvo and is now an industry consultant, worked on the early SAE/TMC committees. As he puts it, "The J1939 CAN is as meaningful to the heavy truck industry today as the web and Internet is to the world. Powertrain performance, emissions and safety would have never gotten to the point that they are right now without the J1939 data link. My highest concern about moving forward is that J1939 is an open architecture, and as an industry we have got to address cybersecurity as we move forward."
Kevin Otto, who retired from Cummins late in 2015 and currently is an industry consultant, adds: “The original J1587 allowed generic scan tools to read the datalink diagnostic and fault codes; however, it was not designed to be effective for speed and control between components. Hence the evolution of J1587 to today’s J1939 with higher speed and a larger bandwidth.
“I feel that the public diagnostic information available [through] J1939 generally meets the minimum criteria for repair and maintenance. Providing for efficient diagnosis and repair using industry-standard tools is a key feature for manufacturers to provide for users. However, features like module programming and control functions between modules must be managed carefully due to cybersecurity concerns. The SAE J1939 specifications provide a means to accomplish all these features using a common data bus architecture.”
Drivetrain Configurations Have Changed
Since those standards were first put into effect, heavy-duty engine manufacturers have had to adapt to extensively more stringent emission regulations as well as the EPA Phase II Greenhouse Gas and Fuel Economy standards. In combination with this, the need to improve fuel economy has resulted in far more integrated powertrains from a single OEM penetrating the marketplace. This integration is driving the proprietary CAN data buses in the market because of the control features that are managed on the datalink.
Paul Menig, CEO Business Accelerants and former engineering leader at Eaton and Daimler Trucks North America, points out that because of today’s vertical heavy truck drivetrains, the communication standards are not as critical for control as they were 25 years ago with multiple OEM components needing to talk for purposes of both health and control. Therefore, like today with the vertical drivetrain, we have high-speed CAN buses that are proprietary and not public.
Menig points out that as we move forward, “There will always be a need for public datalink standard, primarily because of the repair requirements. However, there will also be a need for a next-generation J1939 of the future because of additional electronics and data – especially from the smart trailers of the future.” (He is chair of the trailer electronic TMC Task forces in S.1 and S.7).
Trailer electronics, he says, add “a whole level of complexity to how, as an industry, we will move forward with standards to support maintenance, connectivity, transfer of optics, etc. It does open up a new level of complexity that we did not face 30 years ago; hence, the work we are doing in the TMC S.1 and S.7.”
And Otto agrees that establishing the standards on trailers is a very big challenge that no existing standard can begin to accommodate.
Even with the 20+ years of standard diagnostics, and all the debates in the industry about the “right to repair,” we still have a long way to go. Tyler Robertson, CEO of Diesel Laptops, points out that, “even with the J1939 datalink providing repair technicians fault codes, the diagnostic repair procedures available to the diesel repair technician for the actual repair process must come from the OEM.
“Sometimes it gets to be cost-prohibitive for multiple licenses for the diagnostic repair processes because just about all independent repair facilities deal with multiple drivetrain components. This will only get worse in the future as we see proprietary and data links for vertical OEM trucks.”
The original 8-pin connector and twisted pair protocols for communicating diagnostic information became outdated quickly. Ron Hall, president, CST Fleet Services, was the chief engineer from Microflex assigned to the original datalink project. He points out: “J1708 was informational and with J1587 lasting until 2008, J1922 was informational plus had static control data, and now J1939 allows for dynamic control information as well as diagnostic fault codes to be passed from component to component…. But now J1939 needs to be completely reengineered. Imagine working on a computer that is 10 years old, much less 30 years old.”
I believe J1939 will survive as a standard and have standard data elements populated by the various ECUs. However, the J1939 will be layered with multiple CAN networks that are proprietary.
Braden Pastalaniec, VP of sales at Uptake, points out that telematics providers are challenged to keep up with decoding and transmitting all of the enriched data that is necessary for detailed preemptive repair insights and AI-based predictive analytics. “Due to the proprietary nature of CAN buses and networks currently on heavy trucks, our requests for consistent J1939 data proves to be a struggle,” he says.
This will be the case moving forward, he adds. “As a provider of AI technology that helps operators and reliability teams move from diagnostic and preventive maintenance to predictive maintenance, we have access to data that is more than sufficient to target the major systems from a predictive standpoint and reduce unplanned downtime. Of course, with the addition of proprietary data, the insights could be more prescriptive, but that's not a blocker today."
As we look to the industry at large to keep the necessary standards within bounds for aftermarket repairs, the coordination between TMC and SAE is more vital than ever. As Paul Menig puts it, “What we are seeing (now) is a confusing mix of non-standard standards.”
Next Generation J1939
How will J1939 evolve? My feeling is that J1939 will survive along with embedded proprietary control protocols, along with a high speed ‘ethernet type’ communication standard in the vehicle.
One thing to keep in mind is that trucks have about a 10-year life cycle and trailers have a 20+ year life cycle. With that in mind, Otto makes an excellent point: “Changing and evolving standards while maintaining some backward (and potentially forward) compatibility is exceptionally difficult.We need to be cautious about forcing unnecessary investment into our customers’ businesses.”
Bryan Hennessy, Kvaser field application engineer and previously J1939 systems engineer at Paccar, is chairman of the SAE J1939 Next Generation Task Force, and points out the following: “In some of the chassis, OEMs have as many as a dozen proprietary CAN buses on the vehicle. High-bandwidth communications is needed for radar and graphical/image data, to be communicated to decision-making controllers and ADAS (Automated Driver Assist Systems) in the powertrain. However, because of J1939 and the critical importance of diagnostic information, I feel that both a controls bus and a high bandwidth bus will be on the vehicle in the future. The SAE J1939 Next Generation Task Force must balance the need for accurate controls data provided by the J1939 CAN bus, but also recognize that J1939 will never serve as a ‘communication standard’ serving the transfer of image-based information to the ADAS within the vehicle.”
Lee Lackey was on the original Microflex team with Ron Hall and is now product manager with Noregon Systems. He has served as chairperson of SAE J1587/J1708/J2497 for the past 18 years. “I think J1939 is no longer going to be the diagnostic language, but will still serve as the low-level command and control language for running the vehicle.”
Have we already outrun our headlights? Ken Calhoun, immediate past chairman of TMC, feels strongly that “as an industry, we are 30 years behind the technology. For example, the data scientist community is getting more and more accurate about predicting component health and pending failures, but it feels like we’re years away from providing the typical maintenance director information upon which he can immediately act.”
Calhoun agrees that J1939 provides diagnostic information for repairs adequately, but accurate, actionable insights are way in the future. “It’s like we tried to go from crawling to running, and skipped the walking and trotting stages.”
There’s no doubt that 30 years ago we never envisioned the complexities and heavy truck that we see today. As Jim LeClaire, Wade & Partners, puts it, “There’s a lot to be proud of and we cannot diminish the phenomenal job that J1939 has accomplished.
“With the open architecture that we started with, the ‘right to repair’ wasn’t as big a concern as it is today with all the proprietary CAN buses. We have done a great job in assimilating the data to standards, but looking now at all the technology emerging in the truck and trailer, we have some challenges ahead as regards what should be standards for maintenance communications in vehicle.”
Personally, I feel it’s naïve for OEMs to discount the need for maintenance of highly advanced technologies as we move forward. Things just break. Together as an industry, we must stress the importance, to both the seasoned technology providers and the newer players in our industry, to get involved with the SAE and TMC.
This will hopefully ensure that 30 years from now, we can look back and state that ‘the new communication standards emerging around 2020……. did serve the industry as well as the original J1708/1587 did in the beginning of this evolution.”
Jon White, who is president of Jon White Inc. as well as part of Wade and Partners has formal education as a theoretical physicist, and began his career for the U.S. Navy. He entered the transportation industry in 1976 working with the early development of VMRS and designing fleet maintenance systems. He was the founder of TMT Software (now a division of TMW), Microflex Inc. (acquired by Eaton and then Qualcomm), and was the co-founder of CST Fleet Services in 2000. Currently he is an independent consultant and teacher in the transportation industry specializing in data analytics / metrics / forecasting with the overall goal of achieving minimal costs and maximum uptime for vehicle asset fleets.