View the January 13, 2021 seminar recording
Jennifer Symoun
Good afternoon or good morning and happy new year! Welcome to the Talking Freight Seminar Series. My name is Jennifer Symoun and I will moderate today's seminar. Today's topic is Current Developments in Connected and Automated Trucks.
Before I go any further, I do want to remind you to call into the teleconference for the best audio quality. If you are listening to the audio over the computer and experience any issues, I am unable to fix them as audio quality will vary based on your network connection, computer, speakers, and other factors. Please also keep in mind if you are calling into the teleconference for the audio, you will need to mute your computer speakers or else you will be hearing your audio over the computer as well.
Today's seminar will last 90 minutes, with 60 minutes allocated for the speakers, and the final 30 minutes for audience Question and Answer. If during the presentations you think of a question, you can type it into the chat area. Please make sure you send your question to "Everyone" and indicate which presenter your question is for. Presenters will be unable to answer your questions during their presentations, but I will start off the question and answer session with the questions typed into the chat box. We will also take questions over the phone if time allows and I will provide instructions on how to do so once we get to that point.
The PowerPoint presentations used during the seminar are available for download from the file download box in the lower right corner of your screen. The presentations will also be available online within the next few weeks, along with a recording and a transcript. I will send a link to the recording in the next day or so and will also notify all attendees once all materials are posted online.
Talking Freight seminars are eligible for 1.5 certification maintenance credits for AICP members. In order to obtain credit for today's seminar, you must have logged in with your first and last name or if you are attending with a group of people you must type your first and last name into the chat box.
Certificates of participation are also available for Talking Freight seminars. These certificates may be used for 1.5 professional development hours if accepted by your licensing agency. To receive a certificate, you will need to fill out a form. Please see the link in the chat box. Certificates will be emailed one week after the seminar. A seminar agenda has been included in the file download box for those who need to submit an agenda to their licensing agency.
Finally, I encourage everyone to please also download the evaluation form from the file share box and submit this form to me after you have filled it out.
Today we'll have three presenters:
Our first presentation will be given by Richard Bishop, of Bishop Consulting. Bishop Consulting provides strategy and partnership development in the domain of intelligent, connected, and automated vehicles, advising automotive manufacturers, suppliers, and tech firms. Current clients include Aeye, Outrider, NODAR, Peloton Technology, Plus.ai, Robotic Research, SteerTech, Transurban, and Rijkswaterstaat. BC advises government agencies around the world as well. Mr. Bishop serves as 2nd Vice Chair of the American Trucking Association's Automated and Electric Truck Study Group. He is also founder and chair of the International Task Force on Vehicle Highway Automation, which recently held its 22nd Annual Meeting. Prior to establishing Bishop Consulting in 1997, Mr. Bishop was Program Manager for Vehicle-Highway Automation at the USDOT Federal Highway Administration.
Richard Bishop
Thank you very much, Jennifer. I'm really pleased to be a part of this event, and to be sharing the stage with Hyungjun and Dan. I'm always happy to do things with Federal Highways, having spent some great years in the 90s there at Turner Fairbanks. This is 2021, so I came to Federal Highways in 1991, so this is year 30 in the automation world. I've worked with fabulous people back then and created some lifelong friendships and I'm glad to have stayed in touch. One thing I'll say, I notice on the registration several folks are involved in the transit side of things. Contact me separately, if you like. Even though my talk is about truck automation there are some cool things going on in bus automation too.
My aim is to talk very broadly about truck automation, more from the commercial side, rather than the government side. I will fly through a paramount of things to give you that quick look. A huge part of everything in our lives is about the pandemic. That certainly has pervaded things in the tech community. I would say in automated trucking it has been a mixed bag, it really depends on individual companies, some have been affected more than others. Most, but not all, end-users have retrenched. And generally, by end-users I mean the truck fleets. They're trying to just do the basics. Most have deferred innovation, but some of the larger players have continued on. If you are a startup it depends on where you were in your funding cycle. If you just got a big load of funding in January, 2020, then you are probably in good shape. But the pandemic hit others where they needed to replenish their funds, and that has been challenging. And a new product introduction in this climate is challenging as well, unless you are like Waymo and you're only just putting your own robotaxis on the road. Things have been hampered because of the need to distance people. For on-road testing, it depends on the state and their COVID protocol in that sense. So, it is a mixed bag.
What I will talk about, at a very broad level is, what were the macro trends in 2020. It is important to separate automation, generally, from privately owned vehicles versus fleet vehicles. The auto OEMs are developing autonomy for personal use, but they're not in a huge hurry for that. It's just about a competitive process, and they will do that at their own place. Fleet operations are where the action is; the robo-taxis, robo-delivery. And, if you're not aware, fully driverless services have been launched by Waymo, robo-taxis in Phoenix, and that is starting to spread. It will probably spread slowly, but that is a huge milestone for road transportation generally. We will see how it goes from here.
There is certainly strong momentum in driverless goods movement. So, just a little bit more about the fleet side. Whether it is for robo-delivery, robo-taxi, robo-truck, fleets are a great way to deal with automation development and deployment. Because, as a fleet, you can select your operations area for the best fit with your tech capability. Maybe you're not set up to for the snow; ok, then you run in the southern tier of the USA. That is very different from the OEM world where if you put a truck on the market it is expected to work anywhere in North America. This is a big difference, because these are startups as well. It is a different world than OEMs. The cost pressure is not like the retail model where you have to worry about $.10 on an automotive part; it is all about return on investment. And, of course, you have the option to have hands-on skilled staff to coddle these systems in the early days; software upgrades, system certification, specialized maintenance personnel, that kind of thing. So, that is why this is such an important area.
So, what I am going to do is talk about four sectors of goods movement: controlled environments, streets, resource roads, and highway. I will spend most of my time on highway.
The street world is about delivery of small parcels. Obviously, streets are a complex environment; typically, low speed, a lot of it is retail, so it is customer facing. COVID and the pandemic clearly created a stronger, much more dynamic market in delivery, generally, and it has helped the companies that work in this space in autonomy. So, a couple of quick flash looks at who is doing what. There is a business-to-business side. Waymo is working with UPS to move parcels from the UPS retail stores to their sorting center. A company called Gatik that has just announced customers Walmart and Loblaw. And their most recent press just last month is that they are launching fully driverless operations in Arkansas with Walmart this year. That again is huge. I think we will see 2021 as the year when these fully driverless operations start to happen beyond more than one or two companies. You see they have raised some funding. There is a company out of Sweden called Einride. They've also announced some customers and had success with some funding. And on the business-to-consumer side, Nuro is the most well-known company here, working with Kroger, CVS, Fry's food with a custom-built vehicle that can bring your groceries to your door. So, that is the street side of things.
What about getting to the larger truck world of controlled environments, logistics yards, semi-controlled industrial roads or spaces, such as this port area. There we have low speeds, it is dirty and dusty, very specialized. Because it is so specialized, the truck OEMs have not served this world as much and specialized OEMs have to deliver the typical trucks, . of this type.. It is a small market and therefore, the investment for automation has not happened prior to the current wave of automation development. A company I will focus on here is called Outrider. They are a company founded in 2018. They have several customers and they have announced one, which is Georgia-Pacific. They just raised some new funding a couple of months ago and it is all about automating this capability. A distribution yard or a logistics yard is ideal. It is private property, the environments are well defined, the complexity is constrained, and it is a repetitive task. If you look in this photo, you will see an apparatus that's in blue between the tractor and the trailer, that is a robotic arm that can connect and disconnect the brake line to the trailers. That is absolutely essential. It is one thing to develop an automated truck, but you have to also work that piece if you're really going to have a driverless operation. Jennifer, this is like where I would like to show the video please. This would be Outrider trucks giving you a sense of how they will operate in a typical logistics yard. It's pretty short.
Jennifer Symoun
I will bring that up. Just to let everyone know there is no audio other than the sounds of the trucks moving.
Richard Bishop
Yeah, the audio doesn't matter. There is kind of a control center that monitors what is going on. The electric trucks is a key part of Outrider's approach, so this is an electric automated truck. The shippers who run these distribution yards have a lot of advantages by going electric. They don't have to deal with diesel on the premises, for instance, they can reduce emissions, etc. And there it is. When automation works well, it is boring and that is what it looks like.
So, I will continue to resource roads. This hasn't been talked about as much lately. By the way, I'm writing these days for Forbes.com. I have a link to my Forbes page on the very last slide. I've written articles about most of what I'm saying here so if you want to go deeper you can look there. Resource roads in terms of mining or forestry or whatever, they are public roads, officially, but they're pretty much empty of anything except trucks. They are unpaved, operating in remote areas, etc. And yes, there is a need here for automation as well. I have worked with a group called FPInnovations that's a non-profit research center in Canada serving the forest industry. They have a pain point in driver shortage. No surprise, right? So, they are interested in automated follower platooning so one truck driver can handle two trucks worth of cut timber. They are planning to launch a project this year to work with commercial players and adapt the commercial systems for rural public roads.
Then we get into where the majority of action is happening on the highway. It is a well ordered environment. It's also high speed, so the risk is a large in terms of kinetic energy involved. The market is huge.
There are two primary operational modes: platooning levels and solo driverless. First platooning, there are two generations out there. The most mature is a first-generation level one platooning where there are drivers in both trucks. The front driver drives normally and the rear driver steers and monitors the road. It's just the foot pedals that are automated. This provides fuel savings. Second-generation, is to go to what I call auto follow. There is no driver in the follower truck, and the front driver is providing the human intelligence. Who are the players in this space? The top row are startups and small companies; Peloton, Locomation, and Robotic Research. I have some OEMs at the bottom. They have all been involved in prototyping and testing these. It is unclear whether they will bring the product to market or not but that remains to be seen. They have not made definitive statements about their future plans for platooning, especially second generation platooning.
If you're not familiar with truck platooning, these are the basics. Truck platooning is all about drafting, one truck behind another, which is not a new concept, but to do that with vehicle to vehicle communication, so that you can safely get close to another truck. The fuel economy benefits are really huge; about 4% for the leader, 10% for the follower, at a 60 foot gap and 60 miles per hour. So, not bad at all. The commercial systems focus is on multi-lane divided, limited access highways. Many people ask how will these trucks handle cut-ins by other vehicles? The short answer is, it is a key part of the safety case, the engineering is there to handle that. Now, the following distance laws vary across in the states. So, platooning could potentially violate those laws. However, for two or three years now, the various players have been working with the states to align those regulations with the platooning. This dark green that you see is where a commercial deployment of platooning is now fully allowed. That is a lot of geography. It is over 80% of annual U.S. freight truck traffic and it is certainly enough to start the market here. One example of one of the players is Locomation. They have been in the news lately. T hey have a relationship with Wilson logistics out west, that I would say is a medium-sized carrier. They did some testing with Wilson Logistics last summer out west. And Wilson Logistics has taken a further step to place an order for 1,100 of their trucks to be equipped with Locomation technology. That's a big deal in the truck automation world. Initial delivery is expected to start in 2022.
And then there is another flavor of platooning called leader-follower, in the Army. This isn't that well known outside of military circles. It is a different functional specification. In some situations, they want the typical short gaps in fuel economy savings, but they also want large gaps in hostile environments, for instance. So, the spec here is level 1 and level 4 platooning, meaning driverless. One hundred systems have already been built by Robotic Research under an Army contract. They're being evaluated at an Army base. And the word is that the Army is proceeding with a large buy of thousands of systems with that RFP expected this year. This could affect the entire space if this moves forward.
Now I am finished with platooning, I will switch to solo driverless players. You can see this is a very crowded slide. The top half, more or less, are startups working in this space. It is a very hot space these days. You can also see that the OEM's came along a little later, is how it worked. The startups kind of broke ground and then the OEMs saw the potential and have come along. There is a lot of money going into this space, both venture-capital and internal programs by these truck OEMs. The various players are looking at two modes of operation. One is called ramp-to-ramp, simply meaning that the truck only operates on interstate highways. The simplest environment possible, in this sense. Others are also taking the load off the interstate highway and all the way to the loading dock. The ramp-to-ramp idea involves transfer yards of some sort, immediately adjacent, as you the from that UPS example. This is the Florida Turnpike. So, anyway, dock-to-dock and ramp-to-ramp, those are the discriminators.
I want to give you a few tidbits of what is happening out there. All of these startups are in test mode, they're operating level 4 intent systems on public roads backed up by safety drivers. Level 4 means the driverless truck; no human involvement is needed. They're designed for that and they're backed up by safety drivers as they work their phases and perfect their systems. So, a couple of specific companies. Kodiak is actually doing a regular 200 mile run between Dallas and Houston. These are commercial deliveries. They just announced this week, 800 miles of continuous disengagement free driving over multiple runs. Meaning the safety drivers needed to do nothing on a series of these 200 mile runs. They call that a big milestone. I think others may have similar capabilities, but that is a good example of the maturity we have. That doesn't mean we are there, but it is an example of progress. TuSimple is also running commercial trips each day. They're saying they will have their first driver-out operations this year. They will do this by retrofitting trucks with the needed redundancy as part of the requirement for automation, redundancy of steering, redundancy of brakes, etc. And then in 2024, they're saying there will be factory built level 4 trucks (they're working with the Navistar on this) with that redundancy built in at the factory. They're saying they would then go nationwide with the driverless commercial operations. Waymo, that's the company that spun out of the Google self-driving program, is highly active. Waymo is mainly in the news for Robo taxi, but they also have a strong truck program, and they are sourcing their trucks from Daimler. And then we have Plus. I will say a little bit more about Plus, it is a company I'm working with. They have done pilots with a variety of private fleets; they are testing across many states. And actually, their work includes winter testing with MnRoad. The Transportation Research Center in Ohio is doing independent track-based validation for them. Most of these companies have put out voluntary safety self-assessment reports. You can find those on the NHTSA website. My bottom bullet, several developers note significant fuel savings from automated driving. And that is because, in essence, you have the most perfect truck driver you can imagine, always doing the driving rather than the variation you see among human truck drivers. And then, as I already mentioned, the truck manufacturers are doing things as well. Very important programs there.
I want to highlight an approach that Plus is using in terms of validating these systems for safety. It is a statistical approach. They say that billions of real road miles are needed to statistically prove the safety of the system before making fully driverless trucks commercially available. Many of you are familiar with the Rand report that says something similar that came out a couple of years ago. Well, how do you get to billions of miles? Here is what they are up to. In 2021, they are introducing a product, initially in China; intelligent trucks powered by their level 4 self-driving systems. That means, it is humans involved but the level 4 system is running the truck. This has been jointly developed with the largest heavy truck maker in the world based in China, FAW. So, these will come to market in China this year. Initially it will be a driver-in product, the driver will have to monitor the road. And then incrementally, using all those trucks driving around with a driver in, it is kind of validating that software, incrementally, to allow a transition later to full driverless driver-out product. How you get to 1 billion miles? It is just arithmetic. The typical truck averages 100,000 miles per year or more. Five thousand trucks running for two years at that level is 1 billion miles. Five thousand trucks is not a lot of trucks in terms of the Chinese market. So, this level 4 driverless deployment is expected by about 2024, as the truck OEMs bring these level 4 tractors to market.
Really quickly, I just want to show you the way I see the deployment geography. This relates to both the business case and regulation. The U.S. looks best in terms of the structure of the truck industry as well as the regulatory environment, which is very open. I don't have the time to fully talk you through this. But the reason you see Japan and Europe down at the lower left looking slower is simply because they use the type approval process, which is a very step-by-step process that involves a lot of players. It is not designed to be fast; it is designed to be very thoughtful. It doesn't mean they're doing something wrong, but it is just that in terms of the way that truck automation will roll out, it looks like the U.S. and China will be first.
I will finish up by asking this question, what does automated freight mean to sustainable supply chains? Plus actually ran a load of butter for Land O'Lakes across country from Pennsylvania to California, 2800 miles, last year. And they did this in only three days. Because, although they were using safety drivers, they were emulating the idea of a driverless truck, which could get across country in three days. The typical driver, because of hours of service, takes five to six days. What does this mean? They did see the kind of fuel savings I mentioned before, 15-20%, in this case. And when you think about supply chain you can think about that refrigeration unit. It only had to run for three days for the load of butter instead of six days. Those are the sorts of things that come into play. Can driverless trucks become a new mode for airfreight for time sensitive shipments? That could indeed be the case when you cut the average time in half. I will finish with this. I have only scratched the surface of what supply-chain means and to what automation means to the world of goods moving around. So, my best example is ripe tomatoes in the grocery store. Typically, you see half ripe tomatoes in the grocery store, because they have to pick these tomatoes in the field in California mostly green and then put them in a truck and ship them 6 or 7 days before they get to the grocery store. But with these trucks moving continuously across the country, maybe we will actually have ripe tomatoes in the wintertime in our grocery stores. With that, I will close. Thank you.
Jennifer Symoun
Thank you, Richard. Our next presentation will be given by Hyungjun Park, highway research engineer in the Office of Operations Research and Development of the Federal Highway Administration. As the Technical Program Manager of the CARMA Freight program of the U.S. Department of Transportation, Dr. Park manages various Cooperative Driving Automation activities focusing on commercial motor vehicles and freight. He is also leading the Truck Platooning Early Deployment Assessment project.
Hyungjun Park
Good afternoon or good morning, everyone, depending on where you are. My name is Hyungjun Park, as introduced. And first of all, I would like to thank you for this opportunity today to present truck platooning work that the USDOT has been sponsoring for years. This is one of the high profile projects within the department. Our first speaker today, Richard Bishop, has done a great job in painting the overall picture of the connected automated trucks area by touching on everything that is out there. With that, now I would like to change the gear to one specific project that the USDOT has been sponsoring since 2018. And it is the Truck Platooning Early Deployment Assessment Project.
Just a brief overview of what I will be presenting today. I am going to start the presentation with a brief background focusing on the phased approach that we took. And then I will share with you all what happened in phase one and what has been happening in phase two, implying that phase one has been already completed and we are in phase two right now. For phase one, I will spend only a few minutes to briefly touch on the awardees and several key deliverables from phase one. Then I will move on to phase two and go a little deeper to provide a little more detailed overview, including the project team and the truck platooning system that is being used and the experimental design. And finally, next steps.
Now moving over to the project background. This is the slide I usually use to describe what truck platooning is. But since Richard already provided the basis of truck platooning, I would like to save some time here. However, one thing I would like to point out here is that the truck platooning system being used for this project is level 1, meaning only the longitudinal movement will be automatically controlled by the system, but the driver still has to steer the truck for the lateral movement.
For this project, initially we had several goals and objectives in mind that we wanted to accomplish, and the first goal was to understand the impact of truck platooning operations in a real world environment. In other words, we wanted to see the impact or benefit of platooning real trucks driven by real drivers, for real business deliveries. This goal is really important in that, until this project took off in 2018, there had been probably no activity looking at the extended period of time for the actual field operational test. Most of the activity, until that time, had been targeting one-time demonstration or a limited proof of concept type testing to collect a limited data. But this project is aiming to conduct a field operational test for at least one full year so we can collect the data and do the analysis. Then, the second goal of this project is to assess the benefits and impact of truck platooning operations in terms of various areas of interest such as mobility, safety, environment, infrastructure, and all other various topic areas. We also acknowledge that there are some recent activities like the ones presented by Richard that are mainly initiated by private companies. They are doing similar operational tests of either platoon or solo automation. But one thing that makes this project different from those is that our focus is not only on the platooning trucks or the automated trucks, but also on the overall transportation system. So, in other words, we see the impact on platooning trucks and the impact on surrounding traffic equally important. Finally, based on the result gathered from this project, our hope is to provide valuable information for our stakeholders to assist them in planning and making decisions in relation to truck automation.
This is a high level overview of the truck platooning project. As I mentioned, to accomplish these goals we decided to take the phased approach to minimize the risks due to uncertainties in dynamic environments. So, basically phase one was conducted in 2018 and that is now complete. And in this phase, the main goal was to develop a proposal for field operational tests that are going to happen in the next phase two. I will get back to this later on with a little more detail. At the end of phase one we received several proposals for phase two. We decided to go with one contractor for phase two and the goal of phase two is to complete all of the plans and get the truck platooning systems ready for the field operational test. And then finally, conduct the field operational test for one full year.
Moving on to phase one. In phase one, we selected three teams for the contract that include Battelle, CDM Smith, and California PATH. Their role was to basically develop the proposal for phase two. And for that they have done some fundamental work that is needed to prepare the proposal. These teams were supported by the independent evaluator sponsored by USDOT. now I am going to go over the deliverables one by one to provide a little bit more detail. We have 4 or 5 deliverables from phase one. The first deliverable we received was the Deployment Operational Concept document. This is basically a high-level concept of operations. This document describes the platooning system and the trucks proposed, and it also provides an initial route for field operational tests. Then it defines the operational design domain and the operating rules within those proposed ODDs. And finally, initial set of performance measures developed to meet the minimum requirements were also documented here. Then, the next deliverable is the Test and Performance Evaluation Plan. This is an overarching plan for the whole field operational test. This first defines the performance measures to be calculated, and then identifies the data elements that are needed to calculate those proposed performance measures. Finally, this document provides the methodology on how to collect, manage, and analyze those data sets. Then we had two more relatively short documentations; one is the Partnership Plan and the second is the Phase 2 Readiness Assessment. The partnership plan is a document that identifies the partners for phase two operational tests and defines their roles and responsibilities. More importantly, the commitment letters from those identified partners were included in this plan. Then, there was Phase 2 Readiness Assessment. And the awardees conducted a preliminary readiness test to gauge the feasibility of the phase 2 field operational test. Finally, they incorporated all of their results into one proposal that was reviewed by the government.
Now moving along to phase two. As I mentioned, we had three awardees for phase one, and we decided to go with one team, California PATH, for phase two. The project team consists of California PATH, obviously, and they are going to be the team lead, as well as the technologies provider required for platooning systems. And Roley's Trucking teamed up with California PATH as a fleet operator. This company is a small size trucking company based in Southern California. They are delivering goods from California to Texas on a daily basis, which is really good for the field operational experimental set-up. The project team also includes Westat as the human factors expert and Cambridge Systematics for stakeholder coordination. In addition to these, the PATH team also has some other partners like CalTrans, the California Highway Patrol, and the California Trucking Association. This partnership is still being expanded as we find out more stakeholders or partners that are impacted by or that have an impact on the platooning operations.
This is the truck platooning system. PATH is going to implement the cooperated adaptive cruise control capabilities on 4 new trucks. As I mentioned, the platooning system that will be used in this project is the SAE level 1 automation. It automatically controls longitudinal movement only. They are planning to enhance the existing platooning system that was developed and tested and demonstrated already from previous projects. One of the notable projects is the one that was sponsored by FHWA through the Exploratory Advanced Research Program. Out of four trucks, two to three trucks will be operated in a platoon so we can collect operating data, while the fourth truck will travel closer to the platoon as a control truck so we can collect the baseline data for the same route at the same time. In terms of sensors, what you are looking at on the right-hand side of the slide is a set of sensors, such as lidars, radars, and 3 cameras, and so on.
And this is the experimental design and the proposed route. The proposed route is a 1400 mile segment of Interstate-10 from Rancho Cucamonga, California through Arizona, New Mexico, all the way up to Fort Worth, Texas. This route is mostly rural with a few urban metro areas such as Phoenix, Tucson, and El Paso. This route is mostly flat, but it has a few mountain passes to go through that will give us an opportunity to study the infrastructure or environmental impact on truck platooning operations. In terms of the experimental design, as mentioned before, we will use two to three trucks in a platoon to collect the field operational data and a fourth truck driving closer to the platoon to collect the baseline data. Our current plan is to recruit a total of 20 drivers and divide them into four separate groups with five drivers each. Each group will run the test for about three months so we can have an idea about the variations for a specific driver within the group and across the groups. Finally, our target will be to complete at least one round trip per week. So, if the field operational test continues for a year (this is our plan) you have at least 52 data sets of a 2800-mile trip. This will result in around 145,000 miles driven.
This is the type of data that we are planning to collect. The first set is the engineering data. This is a typical data set like position of the truck, the acceleration and deceleration, and the truck platooning operational mode or something like that. And this data will be collected using typical on-board sensors and through J-1939 bus. The second set of data is about the surrounding traffic. As I mentioned before, we want to look at the impact on the surrounding traffic of the platooning operation. So, we are planning to use extra sensors such as fixed beam lidars and 3 video cameras, one front facing and two side facing that will be installed outside the cabin. Then, there are a couple more sensors that they are planning to use to collect the truck driver data; SmartCap LifeBand and Jungo VuDrive. And finally, we are going to have a wireless modem connection throughout the field operational test to monitor the system in real-time. And that is about the CACC system operation and also about the data logging health.
And one last thing I would like to briefly present. The performance measures is one of the most important things that we were considering in this project. So, what we have done for the performance measures in phase one was to come up with eight categories for performance measures starting from platoon operational characteristics and also safety, mobility, and environment, and also infrastructure impact and other state and local government impact as well. And for each of these categories we have come up with a requirement ranging from 2 up to 12 requirements for each category so that we are not missing anything that is important. Based on these requirements, the awardees of phase one, they proposed a set of performance measures that met those requirements. And there are two reports already published and they are accessible by anybody. This is one report that you can get. If you google the first document, then the second document will come up automatically.
I think this is the last slide that I have about the next steps. So, we have two separate stages in phase two. Now we're in the middle of the Implementation Stage. The project team is updating and completing all the plans. After that there will be a series of testing, including system acceptance, driver acceptance, and operational readiness testing. Then after that, at the end of the implementation stage, the government will make a go/no go decision for the next stage, the Field Operational Test Stage. Our expectation is that this field operational test will start in November of this year and that will run for at least one full year. The actual data collection, evaluation, and all the analysis work will be done during this time. I think that's pretty much what I have for today. Thanks again very much for the time and interest. There are two emails, one for myself and the other is for the team leader, Gene McHale, of the Office of Operations R&D. Please feel free to reach out to us. Thank you.
Jennifer Symoun
Thank you, Hyungjun. Our final presentation will be given by Dan Murray, the senior vice President of the American Transportation Research Institute, also known as ATRI. Dan leads ATRI's substantial portfolio of trucking research activities, is active in TRB activities, and sits on the boards of several university transportation centers. Dan has also led several industry studies focused on autonomous and/or highly automated trucks.
Dan Murray
Thanks, Jennifer. Well, first of all, 10 seconds on ATRI. We are the not for profit research arm in the trucking industry. We do research in a number of different areas. All of our research is available on our website, free of charge. We work closely with both industry and government, we have an ongoing contract with the BTS to run their freight mobility initiative, etc. We also have a board of directors made up of presidents, CEOs, and chairmen, of companies you are certainly familiar with. If there is one in there that you're not as familiar with it is right in the middle there, TransSafe. That is the consulting firm for former FMCSA administrator Annette Sandberg. But they're not on the phone, so I will say more importantly we have a research advisory committee made up of the trucking industry: motor carriers, vendors, we have both teamsters and a wider represented, we have federal government, state government, we have enforcement and academics. These folks have consistently, over the last three years roughly, identified understanding the impact of autonomous technologies on the trucking industry in multiple different ways. We have conducted a bunch of research in this world, both qualitative and quantitative, massive surveys of truck drivers, motor carriers, even Richard Bishop and I were working together on the Auburn University Platooning Task.
But what I'm going to do today is take off my research cap and answer what Jennifer asked me to talk about. Which is, what is the industry perspective on autonomous trucks and Platooning sort of as a subset of that? That is what I'm going to cover today. I'm going to provide the devil's advocacy argument in many instances, because the engineers have done a phenomenal job of hooking up wires and proving that these systems can work, technically, but many of our biggest challenges go far beyond the technical aspects of ATs, autonomous trucks, if I can use that term.
So, first of all, ATRI surveys thousands and thousands of truck drivers and motor carriers every year; we have our Top Industry Issues Report. I highlighted some of the things I'm going to go cover quickly today in orange, but I mistakenly didn't cover in orange Driver Shortage, the number one issue on motor carriers. But you will discover, shortly, that may be the single biggest challenge or problematic issue for autonomous technologies going forward in the future. I will get to that in a short while. This is what I will talk about briefly, again. Almost all nontechnology issues and challenges; ROIs, traffic congestion, e-commerce, litigation and insurance, lack of jurisdiction standards (which is really standardization of regulations and policies, certification), weather, and infrastructure. And what I'm not going to talk about in my presentation, but somebody ought to raise their hand, at least when it comes to connected vehicles, and ask is DSRC versus 5G better or worse? I personally think 5G will put DSRC to bed someday, but that is not a topic I am covering today.
First of all, we have huge ROI issues. The average truck load carrier in the US has an average operating margin of about 3.5-5.5% which is 3.5-5.5 cents on the dollar profit. With the trucks today, a truck tractor costing $135-145,000 and an autonomous tractor costing closer to $200,000, I've got to have an ROI in a relatively short ROI because I turn these trucks over every 3-5 years if you are larger feet. So, the ROIs are not known. How do we know that? Look in the photo in the lower right and you will see almost every single technology vendor testing these systems, even ones saying we've proven level 4, for instance, have at least one driver in the seat, holding the steering wheel, and most have a second attending or engineer in the passenger seat. We have no data, Richard was talking about actual data, but we have no ROI data of what the benefits would be to now spending close to $200,000 on an AT when I still have the labor costs involved, and possibly two labor costs, as they have a requirement in California. Nondriving obligations exist. Federal law today says I still have to do a daily vehicle inspection. I can't do that today without a human and none of the technology vendors have even thought to say, what are the other obligations of the truck driver to see how technology might meet those obligations. All we're doing right now is looking at whether a truck can drive without a human. So, that is not enough of an answer for the industry to say the ROIs are there. Again, expensive equipment. Some folks are even thinking that, certainly at the connected vehicle level, the trailers themselves will have to have transceivers and sensors beyond just the tractor.
Then they get into traffic congestion and e-commerce. Now, platooning I'm not even going to talk about, because in a congested environment, platooning has already failed. I need mainline speeds, I need certain headways, by law. And, of course, to get those 4-10% fuel economy numbers that Richard mentioned, and some people have a hard time getting those, I'm going to need speeds continuous at 55 miles per hour for longer stretches. If I'm platooning across multiple fleets, I'm going to have to alternate between the first truck and the second truck because of the different fuel economies. So, it's extremely problematic that track traffic congestion is growing in the United States. But particularly, when you look at the benefits of these systems and what they can do in an open road environment, most of those benefits disappear. Several of the technology vendors have tested trying to change lanes in a rush hour congestion environment and most of the car drivers won't let you in. It is very problematic, you miss turns, etc. So, traffic congestion is a big issue, and it is probably not one, as you know, that is going away.
Now, I mentioned litigation and insurance which go hand-in-hand. We did some research, released this summer, talking about impact of nuclear verdicts on trucking. And you ask yourself, what in the world does that have to do with autonomous technology? Well, really, it is very clear that 90% of all crashes today have a human factor causation to them. That will probably flip-flop in the future, so when I sue anybody for human failures today, I'm going to sue technology companies, etc. in the future, but at the epicenter of all that is the trucking company. When you look at the impact of litigation and verdicts on the industry over the last decade, you will see these verdicts are substantial. And I tell you this, because in the trucking industry, we are sort of what I'm going to say a is victim of a tort law, which is civil litigation law. In almost every state in the United States, you as a personal driver, but a trucking company in particular, can be substantially less negligent than they are financially liable. Even up to 100% of the crashes when they may be only 40-50% negligent. So, clearly the trucking industry is in the cross hairs of litigation and that drives up insurance costs. In fact, insurance costs today are the third highest cost center in trucking after fuel and labor. It's extremely expensive, but it's not a level playing field. The smaller you are as a fleet, the more you will pay on a per mile basis for insurance. So, this immediately says if the insurance industry which views these systems today as speculative, and we work very closely with insurance industry, there is likely going to be a premium added to your insurance cost for running the speculative technologies. Until again, as we discussed earlier, we have to have billions of miles of actuarial data on safety. And until we get to that point, insurance costs will be very real, and clearly, the smaller you are, the less likely you are to consider adopting autonomous technology.
Which brings us to what Richard was talking about earlier. We released, early last year, a report on the standardization of regulations and policies, certifications, and other requirements in the United States for autonomous trucks. I did not add the maps here; I would urge you to go get the report. But it is sort of a soup of different regulations. So, even when platooning is allowed, the headways may still change. And I guess on paper what that means, if I cross from Minnesota across the St. Croix river into Wisconsin, I have to stop my truck and change some sensor algorithms or something. I'm being a little tongue in cheek, but that's not going to work. Because, as you know, trucking and freight is an interstate sector of the economy, so we have to have broad swaths of standardized policies and regulations and requirements. And today, they are radically different. Even the certification requirements, the number of humans that might be needed in a "driverless truck environment" are all very different. I'm going to take a friendly shot at the U.S. D.O.T., because they've released AV 1.0, 2.0, 3.0. It is a series of guidelines and recommendations. Whereas we need the U.S. D.O.T., and probably even Congress, to step in and develop national requirements and regulations probably under the edict of NHTSA and FMCSA so that we don't have patchwork environments under which interstate trucking will operate. I'm hopeful that that does occur relatively soon, hopefully in the next version, we'll call it AVX, that comes out of U.S. D.O.T.
Next bullet I had on there is weather. Well, MnRoad has an ultimate testing ground here in Minnesota on I-94. Folks have started conversations about how to utilize that. I can't find a single AT vendor willing to come up and operate these systems in January in Minnesota. Because, frankly, the scatter associated with rain and fog and snow whiteouts is bad. But, even from an infrastructure standpoint, major problems exist. For instance, in the lower right you'll see a Minnesota roadway with the state patrol car pulled over. All of these systems use at least video analytics, if not lidar and radar. You have three lines there. One is the paint stripe you're supposed monitoring and following, the other one is a cement seam, and the third is a snowplow line; two will put you in the ditch and one won't. These systems won't know, per se, which roadway to follow, which line to follow. So, northern climates will be relatively challenging for the systems, which is why almost every single operational test you know about is in California, Arizona, Nevada, Florida, and Texas. At some point we are going to have to do this testing in the northern climates so that we can really be comfortable as an industry that the ROI is there as well as the safety.
Infrastructure is a gigantic issue. We can't get an infrastructure bill out of Congress; we haven't raised the Federal Fuel tax since 1993. While I facetiously added some photos here of what infrastructure could look like in a really bad situation, in some ways what might be worse is if we get an infrastructure bill and spend the next 5-6 years and a giant work zone environment. Which, again, would not allow for most AT platooning to occur, which with or without a bill, will be challenging for many of these systems. We are going to have to identify networks with an infrastructure, which from a platooning standpoint includes grade issues, where the infrastructure is adequate to excellent. That would then be the logical network for an AT network. Which, again, could be readily identified, maybe working with ASHTO and others who already grade the infrastructure on the national highway system.
I have about 6-7minutes left. I'm going to talk to you now, since I've been the reality pill of what is wrong, about what is working really well. So, FMCSA has initiated what I believe is a very exciting program; it is the low hanging fruit side of advanced technologies. It is based primarily on Advanced Driver Assistance Systems, which are active safety systems. You will see that we included warning systems in here, but it is primarily active safety systems that provide driver assistance in a huge range of operating environments, whether it is literally congested corridors or open roads. We renamed the program Tech-Celerate Now, or Tech-Celerate, because ADAS is just another one of those buzzwords that leaves you scratching your head. And the team includes the American Trucking Associations, the OOIDA Foundation, Technology and Maintenance Council, and ATRI. And we are going to have a big webinar coming up in early February to talk in great detail about what is behind this. But I did receive permission from FMCSA to share with you some slides on with these advanced safety systems, which are in many ways a precursor to a level 3, 4, and possibly level 5 truck. But they are here today and some of them are even seeing adoption levels increasing. In other instances, the Tech-Celerate program will improve education and awareness and perception and, ultimately, we hope will increase adoption.
So, first of all, these are the bins of technologies we are talking about. You do see the warning systems in there as a bin, but most importantly you see a series of active safety systems (active lane keeping, automated emergency braking). We did put air disc brakes in there as a standalone technology, mostly as a very important precursor to most of these other systems. Disc brakes, while it may be a bit more expensive to adopt and maintain, they are phenomenal in terms of managing the kinetic energy and other aspects of trucks today. So, all of these systems exist today, can be purchased today, although adoption has been relatively slow. Again, because of awareness and perception and education issues. So, we undertook recently with OOIDA a huge survey, hundreds and hundreds of motor carrier executives, hundreds and hundreds, in fact, almost 800 truck drivers to ask them what are the influencing factors, good and bad, when you think about purchasing active safety systems. Which, again, I believe is a solid precursor to what are the issues if you would consider purchasing an autonomous truck. By the way, this is on a scale of 1-7 where the bigger the number the more influential. Someday we will reverse this to make it more intuitive. But the bottom three issues are the largest issues, and they are all financially based. Now, this is a blending of both the motor carriers and drivers, but both have huge obligations and requirements on autonomous technologies and active safety systems to generate a positive ROI with a relatively short payback. Again, you can't depreciate these things over 10 years since many fleets turned trucks over in 3, 4, 5, 6 years. So, cost and value are going to be a big aspect of these systems and it is one of the things we know the very least about. Because, again, technology engineers have not been pressed to answer some of the ROI questions. If you ask, what are the most influential factors on deciding not to purchase these active safety systems, again you will see a lot of pricing related ones. It is very interesting that from a maintenance standpoint, if any of these systems break or are out of calibration, I would not have a clue where to take it. Most diesel engine technicians don't even have the equipment to diagnose lidar and radar, let alone to repair it. So, again, the secondary market to provide maintenance and repair is a big technology purchase cost. But again, I point out you the single biggest issue, which is driver's control is compromised. The truck driver, at the end of the day, is king. The truck driver is not going away; you will not see level 5 until you are well into retirement. So, we have to make sure that the truck driver feels comfortable that he still is in control, that he has the last out, that sort of thing. Even if we design the systems where that is the case, the driver still has to believe. If you ask the carriers what are the biggest issues impacting adoption, in terms of what will motivate me? Again, separate it out again, the top two are cost related. Before you get excited about lower insurance premiums. Commercial insurance is not like State Farm and Liberty Mutual, they don't offer front end discounts like we do when we have airbags and security systems, etc. They require 3-5 years of actual data to see that crashes are reduced before discounts kick in on the premiums. That is going to be a longer-term opportunity. So, now we are left with again purchase cost, tax deductions. But again, driver acceptance is in the driver list. I'm wrapping up soon now to tell you why driver acceptance and driver shortage are so closely intertwined with each other, as well as the success of autonomous trucks.
When you know that the truck drivers are concerned about the technology and are nervous about it being about the number 5 or 6 issue that it is going to put the driver out of business, that it's going to take away their jobs. And the number 1 issue among fleets of all sizes is the driver shortage issue. In a bad economy which is say COVID 2020, the driver shortage issue was still there because of retirements and a series of other externalities. But in a decent to solid economy, it is a terrible, almost crisis-level for the trucking industry. So, the first thing trucking companies are going to do is say, before buying this technology, will the drivers be satisfied with it? Will they accept it? The answer is, according to the driver data, will be no. They're pushing back because they think it would put them out of work. The big blame with that resides with the technology providers who candidly say we are trying to get rid of the driver, we are trying to save labor costs, we're trying to move into level 5. So, the very last thing a trucking industry and a trucking firm will invest in is automated technologies if and when they think it will take the drivers away and scare their drivers away. So, one of the first opportunities for us is to educate the drivers that the driver is going to be sticking around in level 3 and 4 and that these technologies will help you and not harm you. And when we can sell those messages, hopefully through Tech-Celerate program, I think we will be able to leap forward and start thinking about serious adoption, assuming the ROAs are there. Jennifer, I hit my 20 minutes. I will shut up. If you have any questions, you wake up at 2:00 in the morning with some question that wasn't answered today, just shoot me an email. And feel free to peruse all of the research on our website.
Jennifer Symoun
Thank you, Dan. I will start off to the questions typed in. Feel free to keep on typing in questions and we will try to get all of them. Dan, since you just finished up, I will start with you first. I believe this is on the slide about the most influencing factors not to purchase ADAS. The question was, what about security of the load?
Dan Murray
If you're talking about security in terms of cargo theft, that is a huge issue that autonomous technologies don't even address. We have already today, at least for valuable cargoes, trailer sensors, infrared sensors, etc., we have electronic locks on the outside. But cargo theft is probably a hundred billion dollar illicit industry and autonomous technology won't do anything for that. If you're talking national security like cyber vulnerabilities, that is a fair game. All the cyber security issues you hear about in the cars today when they're talking about Ats, basically started with somebody already inserting a software patch into the vehicle. No one yet has basically hacked into a car while it was traveling at freeway speeds without already having the security vulnerability patch in there. So, in the short run, I am not that concerned about cyber security vulnerabilities. With cars, the consequence of the risk is relatively low; you might put one car into the ditch. With a hazmat load in a large truck, it does become considerably more serious. But, again, you would have to know what the cargo is; if it is hazmat or dangerous goods. And then you would still have to hack into the system. We absolutely need to focus on that at some point soon, but it is not one of those risks that I believe, nor does it the industry believe, is right at the cusp of hitting us in the security realm.
Richard Bishop
Richard here. I will agree with that. The industry, going back at least the last 8 years, has been steadily getting better and better on cybersecurity and there are best practices now. Anyone developing automated vehicles has worked that; they have got very high standards. The problem with the cybersecurity question is you can never respond, oh it can't happen, the hack can't happen. Well, that is the world that we live in. But it is very much a part of the engineering in these vehicle systems.
Dan Murray
I would say if you're really interested in that topic, shoot me an email. The American Trucking Associations is set up with the U.S. D.O.T. Freight Cybersecurity ISAC just to address those issues.
Jennifer Symoun
I should mention, for all these questions, any of the presenters can feel free to jump in because I think some of them to probably apply to all of you. Dan, another question for you. In the program, does automatic emergency braking distinguish between systems that detect pedestrians and those that do not?
Dan Murray
That's a great question. Some vendors will say it does, most would say we are working on it. There is a lot successful research data is saying, basically, other vehicles and roadside impediments are well defined and identified, but that smaller items like pedestrians and bicycles might not be. That is a bit of a gray area. Where the truck drivers are concerned is they point out that braking is not always the best last answer to a safety situation. In some instances, accelerating and switching lanes is a better strategy, etc. So, they need to be sure they can have last override over the decision of the automated emergency braking, which is always breaking. Those are fair concerns. But at the end of the day, the truck drivers today, still with these systems always has the last out with decision-making.
Jennifer Symoun
Ok, Dan, another question for you. We will start with you first and then if other presenters want to comment, too. You had e-commerce listed on one of your site. Can you speak to autonomous truck issues in urban areas when it comes to last mile delivery?
Dan Murray
In urban areas versus last mile delivery, most last mile delivery if you're talking classic last mile delivery which is usually for, which is usually a fulfillment center or warehouse to a residential or local business is not being done by class 7 or 8, class 6 sometimes 18 wheeler. So, in pure last mile, you're not going to see anything in the world of platooning. In many instances these systems are not meant for local arterials. When it comes to traffic injection for the 18 wheelers, most of these systems are designed for streamlined routes, open road settings, highways. In fact, almost all of these settings in beta tests are being done on the freeways and highways. Which is why, Richard pointed out, some of his clients are looking at exit to entrance ramp operations, and then literally a human would climb in. I guess, all of these trucks drivers are lined up on the side of the exit ramp, but these humans would climb in and do the last mile to centers and warehouses. I don't foresee a short-term opportunity for autonomous systems in local or urban driving environments, quite frankly. Partially, because of the traffic congestion and partially because of the infrastructure is just radically different from freeways.
Richard Bishop
I will chime in on that one. You are right, our tractor-trailers are not doing that last mile work. But these wily technology guys in Silicon Valley, they keep slicing up the goods movement piece to find a use case that has a good chance, and they can get venture funding for. One of these is Gatik, you can look them up. They are focused exactly that last mile. They're the ones running goods B to B for Walmart in Arkansas. So, that is out there. Whether it will succeed or not in terms of business case is a different question. But it is out there.
Jennifer Symoun
Thank you. Let's see, Dan. Did I hear you say 200 to 300 mile-per-hour on of this slides related to congested? If so, can you explain?
Dan Murray
That will be the ultimate definition of just-in-time delivery if we could hit 200 or 300 miles per hour. I think what I meant is, I mentioned 100 to 200 miles of freeway speeds to maximize platooning. We need, as Richard said, 55 to 60 miles per hour, we need straight roadways, we need low grade, and we don't want impediments, for instance, cars cutting in and out of us at 60 feet; it is going to be likely that they will do that particularly at entrance and exit ramps. Many trucks are banned from the left lanes of freeways in some states, which means I'm going to be in the right lane with exit and entrance ramps. So, if I don't get these long straightaways, those 4-10% fuel economy benefit that Richard described, are going to be substantially below that.
Richard Bishop
You know, Dan and I, we always have a good time when we are speaking together, because I'm always pushing back at him and he is always pushing back at me. So, my take on that is that platooning does not require contiguous runs. It requires being at speed, because that is where the aerodynamics matter. But, for every 1 mile traveled at 60 miles per hour, you are getting a fuel economy benefit. If it is broken up a bit, it is, but on long runs there can be value. Peloton has not named the customers that they have done trials with, but they have done trials with 7 or 8 major customers and published some information on this. They said that cut-ins were quite rare; about once every 600 miles. They said the fuel economy was averaging 7% for the pair of trucks, the fitting with that 4% front truck, 10% rear truck. I don't see that as a big problem, but you have to choose your runs. If you have a highly congested environment, you don't want to do platoons. But there are lots of runs where that is typically not a problem.
Dan Murray
Yeah, and I will just tell you because Richard and I worked on this together, if you go to the Auburn University, which is a U.S. D.O.T. sponsored platooning test we developed the feasibility study for, there's a lot of caveats. You're not going to platoon hazmat, you're not going to platoon successfully in urban areas, you will not platoon across competitors. You know, FedEx and UPS are not going to be trying to help each other improve their ROIs. At the end of the day within a single fleet, on identical lanes, which maybe is an LTL linehaul environment, there are opportunities for sure. But if you're across different fleets you're going to have to debit each other's bank account and keep could switching between the front and back and monitor who had the first hundred miles and who gets the next hundred, because of that delta between the front truck benefit and the back truck benefit.
Richard Bishop
Yeah, it will be within the fleet initially. What is really driving the customers is the hub to hub concept where the driver shows up for work in the morning, takes the load out for 4 or 5 hours, drops the load and comes back, so they go home at night. That is important for retention of drivers. There are a lot of truck operations that way. Jennifer, if you don't mind, there's been some chatter about Daimler and platooning. Can I address that real quick please?
Jennifer Symoun
Yes, please do.
Richard Bishop
So, I wrote back into the chat, but just to be clear, Daimler did say about this time two years ago at CES, that they were stepping back from platooning. Okay, and they also said they were leaving the door open to come back to it if conditions change. I think that was because the business case for platooning in Europe is not great, and in other parts of the world. In the U.S. it is much better, but Daimler had to take a global viewpoint. I am personally convinced, even though Daimler has not said this, that they are working on auto follow platooning. So, you can't just talk about the platooning business case, you have to talk about is it first generation or next generation. That is where it gets messy in these kind of discussions. I think all the trucks OEMs are working it. The truck industry is incredibly diverse. Platooning is not for everybody, but there are enough players out there that I think it does have a role to play. Will it scale up? I can't say how much it will scale up over time, but there is room for all of these different types of operations.
Dan Murray
One of the technologies in our Tech Celerate Now program is electronic mirrors. Believe it or not, at least anecdotal data says, when you remove one those giant mirrors on the side of a truck and replace it with a video mirror, you essentially reduce drag by up to 1-3%. So, now you have a device that you can put on the outside, which has already been approved FMCSA and NHTSA, that will expand your field-of-view from a safety standpoint, but also improve your fuel economy up to 1-3%, just by replacing a mirror with a camera. So, there is another opportunity for you.
Richard Bishop
And then, of course, if you're platooning with those trucks, that is an added fuel benefits.
Jennifer Symoun
Alright, let's see here. Dan this question came in while you were presenting, but again, anybody can respond. How well do tandem trailers compete with automated trucks?
Dan Murray
I've been in the industry 26 years and I'm not 100% sure I know what they're talking about. By tandem trailers I'm assuming they're talking doubles, and I think what they're doing is how well do automated technologies work with LCVs, longer combination vehicles and doubles. Richard, I'm going to go out on a limb and say the technology vendors I have worked with have never tested anything other than a 53 foot trailer when it comes to combos. I'm not aware of double and triple tests out there.
Richard Bishop
They have tested with doubles, because customers requested that. It's a little more engineering but it is not that big of a deal. I think the question probably comes from, well yeah, you can do platooning, but why not just do double trailers, then you don't have to go through all that trouble with technology. That just really depends on the nature of a trucker's operations.
Jennifer Symoun
Alright, thank you. Hyungjun, there is a question for you, and I know you typed in an answer. I think we have quite a few people here who may just be on the phone since there were issues with logins for this one. So, I'm going to ask it again. Is UL4600 accepted as homologation requirement for autonomous or automated trucks?
Hyungjun Park
My short answer to that questions is no. What I know about that standard is that the underwriter laboratory has been developing UL4600, which is the standard for, if I remember the name correctly, it is the standard for safety evaluation of autonomous products, in general. Later part of last year, Underwriters Laboratory decided to a spinoff from that 4600, and they decided to establish a new standard, which is UL 4600-2 focusing on trucks. And USDOT is part of that. As far as I know, they are now in the process of creating the standards technical panel for that new standard, so nothing has been done yet. But given that a few folks from the USDOT are part of the technical panel, once that standard is completed, will have some impact. But I am not sure if that will be accepted by the government.
Jennifer Symoun
Thank you. Richard, will go back up to some questions for you starting with, what is the infrastructure build requirements to enable this automation?
Richard Bishop
Basically, nothing. There are ways the infrastructure can help, but first I say nothing, because you can't get any funding from venture capitalists if your company has to wait on highways to be upgraded. You have to be able to operate on highways as they are now. If we could have retrofitted the highways with a bunch of helpful things 10 years ago, we would've had this a lot sooner. So, for highways in particular, it is more complicated on the streets. It is really about the lane markings, the standard markings that are out there, bringing them up to the top standards that have already been established. I would say that is the thing I hear most. There is also a very specific item that is important. At a typical exit point off of the highway, the lane marking on the right disappears, there is no marking there. Some are discussing putting a dashed line across that area, and the same for the merge point beyond the interchange, because that gives the cameras a little help in confirming that it is proper to keep going straight. I don't know of a formal process looking at that, but I've heard Federal Highways and some state D.O.T. players talking about that for highly automated vehicles. Work zone data; I'm really glad Federal Highways is working the work zones data question very tightly, very strongly. That is important. Again, the AV players that I know of, they can handle just about any work zone. Kodiak AI just put up some video on their Houston to Dallas run showing them operating through work zones that were not that complex. But, anyway, handling the basics. But there are a lot of complex work zones out there, as we know. It would be great to have work zone data being uploaded to the cloud just to help.
Dan Murray
Yeah, this is Dan. More than one 1 of 4 fatalities in a work zone involve a large truck. So, again, that should be one of our top priorities given that we are on the cusp of a new infrastructure bill to make sure we can handle work zones. Because right now, without autonomous technologies, it is a huge safety issue for the industry.
Richard Bishop
Yeah, and I was just going to speak to the surface streets. It would be helpful have information coming off the traffic signals, the information that is discussed, low latency, DSRC. But even if that was uploaded to the clouds, it may not be low latency, but it would still help trucks approaching those intersections. Again, the full safety case is made based on the equipment on the truck. It cannot depend on any information coming from off board. At the same time, information, like the traffic signals, is just another layer of certitude about the state of the traffic signal. It would be welcomed, but it is not required.
Jennifer Symoun
Alright, thank you. We are down to about 5 minutes left, so let's try get to these next once quickly. If we need to stay on for a few extra minutes, we can do that. Does FHWA have a position on whether or not these projects are eligible for federal funding of automation support projects in MPO areas?
Dan Murray
I can't answer that other than to say that the Auburn project that Richard and I were on was a U.S. D.O.T. funded Exploratory Advanced Research grant. So, is there D.O.T. funding to support testing and field testing evaluation? Yes. If, you're talking formula funds for states, my inclination is no. But there are a bunch of people registered for this that probably could answer that.
Richard Bishop
Yeah, you know, the things I talked about, they're not projects in a sense of this community. They're private sector operations on the roads with really no involvement by the public sector other than regulatory allowance.
Hyungjun Park
Yeah, as far as I know, a couple of comments have come up to my mind. There is no dedicated funding for this. But there are certainly some opportunities, like ATCMTD grants, which is the Advanced Transportation and Congestion Management Technologies Deployment grant, that is one of them. And, if I remember correctly, a couple of projects that got selected for the next round are about freight automation. So, this is the kind of opportunity we can look for. Other than that, I don't know if there's any dedicated funding for this opportunity.
Jennifer Symoun
Alright. Have there been any risk assessments related to automated trucking operations?
Richard Bishop
Yes. It is done by the private sector players. If you're in the public sector it is frustrating, because you don't get to look at that. It is a self-certification world. The ones I know of are quite responsible and doing it right. But that is the world we live in. NHTSA has an advanced notice for public rule making out now discussing how the federal level might get involved with the safety validation.
Dan Murray
One autonomous truck vendor did approach FMCSA formally asking for an exemption, and it was denied.
Richard Bishop
Right. It was a pretty flaky request, but yes, that is true, it was denied.
Jennifer Symoun
Alright, next question. What are the top three things private companies need to implement AV/EV truck technology, if anything, from public transportation agencies?
Richard Bishop
I will just say, they don't need anything. But Dan, go ahead.
Dan Murray
I was just going to say, infrastructure is huge. But I know all of us at every jurisdiction level, that is contingent on Congress passing a new infrastructure bill. But when that happens, I think we need to make sure that autonomous technology requirements for cars and trucks are built into TIPS and STIPS and anything else at a project level to make sure whatever the needs are of these systems, that that is sort of a di facto inclusion.
Richard Bishop
Yeah, that dialogue should be happening. It is in various places in terms of what might the automated driving systems used, as I referred to before, but it is certainly not a need. Dan mentioned the problems that an automated trucks might have in changing lanes in congestion. I would say that the engineering requirements on these vehicles is to do everything we do on the road now. If they can't do that, nobody is going to hire them to run their freight. So, they are doing that stuff. That is their job and they are maturing their systems, but the basic capability is there now.
Jennifer Symoun
Alright. Do you have any more details and additional resources on the levels of platooning for trucking and how they correlate with automation levels 0-5?
Richard Bishop
I don't think that is been nailed down. It is pretty simple in the sense that first-generation platooning is level 1, because all you have to do is automate the longitudinal control and you get the fuel economy benefits, the drafting. And the auto follow is a level 4 because there is no driver in the rear vehicle, but there is a human involved, so, the SAE levels don't quite fit. I would say really quickly, SAE level 5 has been mentioned a couple of times. There is really no reason to focus on that. That is, simply, you can automate driving in any situation whatsoever, but there may not be a business case in which to have such a complete system. The business cases are all at level 4 and below.
Hyungjun Park
One thing to add quickly is, if you're talking about one truck out of several trucks in a platoon, then the SAE 3016 is the automation level, they go level 0 to a level 5. But, if you're talking about multiple vehicles or multiple trucks in a platoon, they will have to look at the SAE 3216, which is the technology for cooperating automation as well. The SAE 3216 is published and it is free to use. So, if you google that, you will come up with one diagram that has the level 0 through 5 up top and class A through D on the left and that will give you a better idea.
Jennifer Symoun
Thank you. I know we are out of time. If our presenters are able to stay on just a few more minutes we will try to get to the rest quickly. If you do have to go, I completely understand. Let's see here. Richard, this one came in for you. One of the slide mapped platooning by state. What states are allowing driverless platooned vehicles?
Richard Bishop
The driverless followers, it's on my slide for whenever you want to download it, but it has been explicitly allowed in Arizona, Texas, and Utah. I will say also, it's not in my slide, but there are a large number of states that allowed driverless generally, whether cars or trucks, and they allow platooning, generically. So, you can kind of overlay those and say that is also where automated follower is allowed. I would say the ones who are deploying the autopilot platooning, I am sure will check in with the states and get clarity on that, etc.
Dan Murray
One of my soap boxes, and it is a very easy small one, is that we should just nationally ban the word driverless, because almost every single one of them, including Google, has a driver in there. Certainly, an engineer plugged into the computer.
Richard Bishop
I disagree.
Dan Murray
What we should do is focus on self-driving. Because, again, all the way through level 4, you will have somebody in here with a CDL, at least in trucks. At least through level 4, there will be a driver and we should focus on self-driving, mostly among the public, Richard, not so much among us, we get the idea.
Richard Bishop
Okay. The point I want to make is that the business case is for driverless, where there is no CDL holder, there is no engineer in there. When you see those two people in the truck cab, that is for test and development. There is no business case for level 3 at all. So, really, we're talking about either the level 2, the lower-level platooning, or we're talking about truly driverless vehicles.
Dan Murray
Right. And frankly, and I was involved in I think the only true level 4 test that is been done was in Colorado where the driver was sitting in the sleeper for 3 hours on 126 miles. But all the level 4 testing being done on I-10 today has one to two drivers sitting in the driver seat.
Richard Bishop
Yeah, but because it is all pre-product, it is not product yet. The product will not have anyone in the driver seat.
Dan Murray
Don't worry, Richard and I are friends and we buy each other beers, so don't be worried about this.
Richard Murray
You owe me, more than I owe you!
Jennifer Symoun
Alright, another question, Richard, for you. Can you talk more about the distribution application with Outrider? What are the origin and destination point for the trailers and are they just staying in the same yard?
Richard Bishop
They are staying in the same yard, so I should have been clear on that. The idea is, within a single yard, and some yards are quite big and intricate, that is where these vehicles operate. They never go outside of the fence.
Jennifer Symoun
Alright. There are also several options on truck electrification. Do you see any significant linkage between the two?
Richard Bishop
Dan, what do you think?
Dan Murray
Can you say that a different way?
Richard Bishop
Basically EV, does EV interest with AV? That is how I hear it.
Dan Murray
Yeah, where it might be interesting is where we get road electrification, which I'm kind of excited about. At that point I could see AV managing which lanes provide electrification and then using basically technology smarts to manage the electrification. Short of that, I don't see an overlap.
Richard Bishop
Yeah, I would say they're orthogonal. You can automate an electric vehicle if you want, but it really depends on the truck operation.
Jennifer Symoun
I think this is probably the last question we will get to today. Hyungjun, this came in during your presentation, but I think anybody could answer it. Is security being considered in case of a hijacking scenario?
Hyungjun Park
Not in the current platoon project. I know that FMCSA has recently completed a security case for automated trucks, but that is not for platooning trucks. That is something that will be published very soon, but that is not in public domain yet.
Jennifer Symoun
Thank you. I know there are still a few remaining questions that came in the last minute in the chat box. I think we will close it out for today. If anyone wants to stay on a few extra minutes and type in quick responses. I want to ask respect everybody's time for today. If we don't get answers typed into the chat box I will be judge of the presenters and get responses that we can send out to everybody. I do want to thank all three presenters. We have had a very engaged audience. It seems like everybody was interested in this topic today. Thank you for your excellent presentations. Thank you everybody in attendance. I will send out a link to the recording of today's webinar within the next day.
The February talking freight is not yet available for registration but once it is an announcement will be sent through the Freight Planning LISTSERV. The Freight Planning LISTSERV is the primary means of sharing information about upcoming seminars. I also encourage you to join the LISTSERV if you have not already done so.
