Photo courtesy of TuSimple
Technology companies like TuSimple are working together with truck manufacturers to develop and deliver autonomous commercial vehicles.
Technology companies like TuSimple are working together with truck manufacturers to develop and deliver autonomous commercial vehicles.
Technology companies like TuSimple are working together with truck manufacturers to develop and deliver autonomous commercial vehicles.
Technology companies like TuSimple are working together with truck manufacturers to develop and deliver autonomous commercial vehicles.
Technology companies like TuSimple are working together with truck manufacturers to develop and deliver autonomous commercial vehicles.

How active safety systems pave the way to vehicle autonomy

Dec. 3, 2019
The road to autonomous trucks starts with active vehicle safety systems, and will continue with data collection and advancements in system technology.

Active safety systems are nothing new for the commercial trucking industry. However, it seems the industry is reaching a tipping point at which comprehensive and numerous safety systems can be integrated into a fleet, either at the time the vehicle is manufactured or as a retrofit specification, to deliver the benefits they offer and reach new levels of autonomy on a widespread scale.

Currently, the rate and scope at which fleets choose to implement such technologies into their vehicles varies widely. As more fleets adopt these technologies, the database of information they produce grows ever larger, and with this expanse of data comes a clearer understanding of how these systems can benefit a fleet and the industry as a whole.

The current state of safety systems

Safety, perception, control, and automation are among the primary benefits active safety systems can provide when specified on a commercial vehicle. These active safety systems monitor a vehicle’s surroundings to a greater extent than what the human eyes of a driver can accomplish. Control over vehicle maneuverability also goes beyond what a driver – experienced or not – is able to observe and translate into action, as utilization of perception devices and the translation of the gathered data is relayed into maneuvers actuated by the vehicle. Automation is achieved through varying degrees and in varying circumstance; from emergency vehicle braking to driverless vehicles, automation of a vehicle relies on the further integration of components and systems into existing systems such as acceleration, steering, and braking.

Some of today’s most common active safety system features include collision mitigation and emergency braking, active cruise control, and lane keep assistance. An active safety system will first alert the vehicle operator through haptic, visual and/or audible warnings. If the driver does not respond, the system will engage to mitigate a road incident.

Collision mitigation and emergency braking go hand-in-hand; utilizing forward-facing radar and camera technology, objects in front of the vehicle are continuously monitored. If imminent contact with an object, such as a vehicle or pedestrian, is detected, these systems can relay information to corresponding vehicle control systems to initiate deceleration and/or braking.

Adaptive cruise control systems also utilize radar and camera technology to establish and maintain a predetermined distance between other vehicles. Some adaptive cruise control systems can even bring the vehicle to a full stop and then resume driving, in accordance with the surrounding traffic, all without driver intervention.

Active safety systems surrounding lane assistance utilize side-mounted radar systems, as well as forward-facing camera and radar components in order to continuously monitor the surroundings of the entire truck and trailer. Features include blind spot monitoring and corrective steering.

Specifying vehicles with active safety system technology is ongoing. Continued adoption of these active safety systems will allow fleets to create, gather, and utilize data sets from these systems. It remains to be seen how the data these systems create and collect will be be utilized by fleets, truck original equipment manufacturers (OEMs), and technology companies in order to continue down the path of more wide-spread vehicle automation.

From active to autonomous

Active safety systems can automate select vehicle procedures and maneuvers for drivers. Detection, perception, and mitigation systems, in tandem with actuation of vehicle maneuvers beyond a driver’s direct influence, is a form of automation.

The Society of Automotive Engineers (SAE) has created a reference table that provides definitions for the six levels of vehicle automation. Level 0 refers to no automation, while Level 5 automation means the vehicle can operate in any conditions without a driver. The more advanced active safety systems are – which is determined by the levels of detection and perception available within those systems, and the ability to accurately translate and transfer data into vehicle maneuvers – determines the SAE level of automation for a specific vehicle.

Integration of the components utilized in active safety systems can lead to the automation of more vehicle systems. A few companies have been working to test and further develop these technologies.

TuSimple is a technology company working with truck OEMs to develop and deliver autonomous commercial trucks operating on the road today.

“We are bringing a new technology, which is autonomous systems, to maneuver the vehicle in terms of lateral and longitudinal, and we have to see and think through the vehicle,” says Vivian Sun, head of business development at TuSimple.

TuSimple currently has about 40 autonomous trucks operating in the U.S. The company has goals of scaling to hundreds of thousands of trucks in the future. Sun says this will be realized by developing autonomous trucks at the manufacturing level while partnering with OEMs to design a truck specifically built to operate autonomously.

“We started a relationship with Paccar in 2017 and we started working with Navistar in 2018 … working collaboratively with OEMs to have an integrated solution from the manufacturing level has been our strategy,” Sun says. “We believe that this technology and the autonomous system needs fundamental changes on the vehicle chassis and all component levels of the trucks, so we have a very integrated approach with the truck OEMs.”

With this collaborative OEM integration approach, fleets would be able to integrate these vehicles into an existing fleet and still collaborate within their networks, maintenance practices, and warranty procedures.

Similar to active safety systems, vehicle automation is achieved through the addition of sensors and computing power, further enhancing the vehicle’s perception and performance capabilities.

Sun says that TuSimple’s autonomous system can be categorized into three layers: see, think, act. To “see,” or observe, is achieved through sensor technology with cameras, radar, and lidar, as well as global positioning systems (GPS) and inertial measurement units (IMU) that TuSimple has been testing. This first layer of the system perceives the environment and delivers a representation of the vehicle’s surroundings.

To address the second layer of TuSimple’s autonomous approach, “thinking,” the company partnered with visual processing system provider NVIDIA to use their engine control unit (ECU) platform for calculating captured sensory data.

“After we have observed all the information from the road, [we have to] process the information and make calculations of all the vehicle’s information around it,” Sun says. This information is processed into action.

The last layer of the TuSimple autonomous process, “act,” refers to sending the translated information throughout the vehicle. Once the information has been perceived, then calculated, it is delivered to the corresponding vehicle components to initiate vehicle system response, from steering, to braking, to activating turn signals.

Global truck manufacturer Volvo Trucks also has an autonomous vehicle in operation for certain use cases, named Vera. Vera is a cab-less vehicle, embedded with cameras, lidar, radar, and ultrasonic sensors. Vera is controlled and monitored via a separate control center. Vera’s extensive sensory equipment “ensures it can handle traffic situations encountered along its intended route in a safe manner,” says Johan Larsson, director of autonomous solutions, Volvo Trucks.

Vera’s first assignment is a low speed, hub-to-hub application, Larsson says. Volvo Trucks is further developing technology, operations management, and infrastructure in order to deliver autonomous solutions for highway applications.

Commercial vehicle manufacturer Daimler Trucks North America (DTNA) has worked on autonomous truck technology with the development of the Freightliner Inspiration Truck. Developed from a Freightliner Cascadia Evolution, the Inspiration Truck serves as a proof of concept for the company’s Detroit Assurance 5.0 suite of safety systems. Active safety systems such as the company’s proprietary Adaptive Cruise Control PLUS with adaptive cruise control down to 0 mph, Active Lane Assist lane departure mitigation, and Highway Pilot Technology were integrated into the Cascadia Evolution to transform the vehicle into the autonomous Inspiration.

Highway Pilot Technology “links together a sophisticated set of camera technology and radar systems with lane stability, collision avoidance, speed control, braking, steering, and other monitoring systems,” says Brian Daniels, manager, Detroit powertrain and component product marketing, Daimler Trucks North America. “This combination created an SAE Level 2 autonomous vehicle operation system.”

Maintenance implications

Maintaining the active safety systems themselves, the components of these systems, and continuing to maintain the entire vehicle will require a new dynamic approach for maintenance personnel as active safety systems become further integrated into commercial vehicles.

A common maintenance practice a fleet may already be familiar with is cleaning sensors. With many radar sensors being installed in the center of the front bumper, and cameras installed behind the windshield near the top center, maintenance includes simply keeping the bumper clean and clear of debris, as well as maintaining windshield visibility, says Ognen Stojanovski, chief operating officer and co-founder of Pronto. Pronto develops highway safety systems for commercial trucks.

What about the components themselves? DTNA’s Daniels says that Detroit Assurance system does not require any periodic maintenance, and that the radar it uses is self-calibrating.

Bendix Commercial Vehicle Systems’ sensors also require little to no routine maintenance, says TJ Thomas, director of marketing and customer solutions – control group for the company. However, the energy management solutions supplier does provide its safety system’s Service Data Sheets online at bendix.com to deliver an explanation of trouble codes, and to detail necessary diagnostic tools.

If maintenance is required for these systems, fleets will most likely need to work at the dealer level for service. For instance, most Volvo Trucks dealerships are certified to handle maintenance for Volvo Active Driver Assist (VADA), a comprehensive collision mitigation system, says Ash Makki, product marketing manager, connectivity for Volvo. Volvo offers certification through Volvo Trucks Academy, which requires in-person training, where technicians can become certified to handle the specific software and parameter updates associated with VADA.

Pronto does not require fleets to update the company’s Copilot software, Stojanovski says, as the company maintains the software. Copilot is a retrofit highway safety system for commercial trucks.

“I would say the challenge for the technician might lie in understanding the holisticness of the system, and how the system work[s] from one component to another,” says TuSimple’s Sun on maintaining autonomous vehicles. “We have a training program with a community college in Tucson, Arizona, where we test our vehicles. We have created a curriculum that is designed for truck drivers to learn to be autonomous truck operators. Similarly, we are thinking [about] what we can do for fleet technicians. Many things are going into the manual with training so that the fleet technicians know how to deal with and interact with autonomous driving systems.”

Active safety systems and autonomous systems will impact routine maintenance schedules for fleets that integrate these technologies into their operations. This will require a re-evaluation on service intervals as operational implications with these technological systems are not yet known.

Consider an autonomous truck with the ability to operate 24/7; how might that impact braking system service intervals? Hours of operation, say at night with fewer vehicles on the road, as well as the calculated actuation of automated braking systems could lead to less hard stops. Conversely, unbeholden to driver hours of service regulations and operating for extended hours could impact the frequency of brake system service intervals. 

To understand these maintenance implications, OEMs and technology companies offering such systems will have to continuously collect use-case data from fleets that specify their technology. This data can construct realistic expectations for intervals of routine maintenance and possible adjustments to warranties. 

Bridging the gap

Maintenance challenges aside, how will the industry bridge the gap from the specification and adoption of active safety systems to autonomous vehicles implemented nationwide at mass scale?

“Active safety systems are the basic prerequisite for a safe autonomous solution,” says Volvo’s Larsson. “Today there are limitations in the technological capabilities, but we foresee rapid development in many areas.”

The major technological hurdles that need to be addressed include utilizing artificial intelligence to enhance machine learning and predictive analytics, as well as building self-sufficient, redundant operating systems.

“Why isn’t anyone [at] Level 4 [SAE autonomy]?” Stojanovski asks. “Fundamentally, we are missing a scientific breakthrough on the artificial intelligence side that we have not achieved [as an industry].”

While the computer processing integral to active safety systems is adept at recognizing and understanding sensory data, these systems have not yet been able to predict potential and imminent environmental changes to the speed and accuracy with which a driver can.

A challenge with environmental recognition, without adequate machine learning, is that the autonomous systems’ functionality is limited to recognizing the environments in which it was developed and tested.

“We have been testing in Arizona, New Mexico, Texas, and southern states along the I-10 corridor,” Sun says. “We do have the ambition to provide autonomous driving service nationwide in the coming years … [the challenge is that] states have different weather conditions, different road construction, different networks. Those are basically the details we have to go into while we are trying to roll out nationwide.”

TuSimple continues to test and develop its system under extreme weather conditions, as well as tackle issues like night driving, in places outside of their current areas of operation, in order to increase the versatility and adaptability of their autonomous system.

Vehicles operating at SAE Level 3, Level 4, or Level 5 will require system redundancies that aren’t necessary at the lower levels, says Richard Beyer, vice president, engineering and research and development, Bendix.

“Beyond Level 3, systems will need redundant sensor hardware,” Beyer says. “In addition, the vehicle will require redundant braking systems and steering systems, all with redundant power supply feeding those redundant systems. Vehicle architecture will need to support the redundancy requirements for Level 3 and beyond systems.”

Conclusion

As the technology that active safety systems and autonomous systems are built on is not new, its specification into vehicles and the software it is integrated with are changing the scope of commercial trucking.

Vehicle safety and efficiency can be realized through active safety systems that are available in the market today. Some autonomous solutions are operational in test markets today as well, though limited to certain applications and locations.

With a shift in operational procedures comes a level of uncertainty. Only time will tell how these systems will change day-to-day operations, as well as the future of the industry. As more fleets adopt such systems, more data will be produced, a picture will be painted, and the progression of technological advancement will be determined by those willing to step onto the uncharted road. 

About the Author

Tyler Fussner | Associate Editor | Fleet Maintenance

Tyler Fussner is Managing Editor - Community Manager at Supply Chain Connect, part of the Design & Engineering Group at Endeavor Business Media.

Previously, Fussner served as the Associate Editor for Fleet Maintenance magazine. As part of Endeavor's Commercial Vehicle Group, his work has been published in FleetOwner magazine, as well as Bulk TransporterRefrigerated Transporter, and Trailer-Body Builders.

Fussner's May 2022 print feature 'The dawn of hydrogen trucks' was named the best single technology article in B2B by the judges of the 2022 Folio: Eddie and Ozzie Awards. Fussner was also awarded Silver in the Technical Article category for the Trade Association Business Publications International (TABPI) 2021 Tabbie Awards.

Fussner previously served as Assistant Editor for Endeavor's Transportation Group on the PTEN, Professional Distributor, and VehicleServicePros.com brands.

Fussner studied professional writing and publishing at the University of Wisconsin-Whitewater. He has experience in shop operations, is a Michelin Certified Tire Technician, and a Michelin Certified Tire Salesperson.