RENO, Nev. — A panel at STN EXPO explored the regulatory, supplier, academic and school district perspectives on implementing autonomous features on today's and tomorrow's school buses.
A fact all the experts agreed on during the July 11 general session was that school buses will never be completely driverless. "You’ll always have someone behind that wheel, if for no other reason than parent satisfaction," said panelist Mark Swackhamer, assistant director of Transportation for Alvin Independent School District in Texas and a nationally recognized fleet manager. Fred Andersky, director of customer solutions and marketing for Bendix Commercial Vehicle Systems, added that the technology for highly touted “driverless” systems is still currently expensive and faulty. Instead, there are various operational features or instances on and around school buses that can be automated to further increase safety while also limiting driver distraction.
That said, Dr. Thomas Barth, a survival faculties investigator for the National Transportation Safety Board, opened the discussion with an overview of recent school bus crashes currently under investigation, including Chattanooga, Tennessee and Baltimore that occured last November. He also revealed the agency's “Most Wanted” list of safety improvements recommended to reduce crash frequency and severity as well as technology in commercial vehicles including school buses to achieve these results. That list included increasing implementation of collision avoidance systems, requiring medical fitness of bus drivers, reducing fatigue-related accidents and strengthening occupant protection.
It’s no secret that new technology doesn’t come to school buses as soon as it is available for passenger vehicles or even commercial trucks. However, the panelists pointed out what is available for buses and the ways to go about best utilizing that. “How can we guarantee safety through intelligence?” asked Professor Kostas Alexis of the University of Nevada, Reno's computer science and engineering department. He is currently working on UNR's "Intelligent Mobility" project along with the Reno Transportation Commission and Proterra to build a zero-emissions, electric and autonomous transit bus.
Andersky spoke about three crucial safety assists currently available for school buses that lend themselves to more automated operations for drivers: Stability systems that work with the ABS systems; collision mitigation technology that works with installed cameras to alert the driver, cut power, or apply brakes as needed; and data that can be used to see what is happening on the road and can be used for driver training purposes, such as lane departure warning and speed sign recognition. He stressed that stability is the foundation for bus safety technology; other features like collision avoidance systems can be added later but you can’t retrofit stability systems.
Swackhamer, who was responsible for bringing GPS and camera systems to school buses in the Texas school districts he's worked for, added that just as important as collecting data is segmenting and accessing specific data, like braking techniques on icy roads. Similar data can then be used to pinpoint unsafe practices that can be reviewed with drivers to increase their aptitude behind the wheel. Andersky discussed existing and developing data transmission capabilities that give real-time reports on driver behavior or provide a feed that goes straight to management’s cell phones, for example, so any recorded unsafe behavior can be immediately addressed with the driver.
Barth questioned how technology can help prevent children from being injured by their own bus, such as in the "danger zone" at bus stops. Andersky said that front-facing cameras are available but that a 360-degree view is better, and existing technology already available for school buses should be integrated with braking. Swackhamer urged attendees to encourage camera vendors to add this functionality.
Andersky explained that electronic braking sensors can help by realizing when a child is near and preventing the bus from striking them. “More automation will help drivers to do their job better and be safer,” he said.
Alexis shared that there could be more autonomous functions available for highway driving that don't translate as well to neighborhood routes. The reason, he added, is technology needs lots of details about the area in which the vehicle is located in order to drive the vehicle and, when driving in close quarters among other motorists or objects, interfacing between the vehicle and its surroundings needs to be reliable. One example of an area that needs further development, he said, is the fusion of capabilities of both radar detection and video cameras to consistently and correctly identify objects, and whether they are moving or stationary.
Barth commented on not only the commercial vehicle technology revolution but also on the evolution of regulations. Andersky brought up the “concerning” exemption of school buses from extra safety measures like stability requirements. Swackhamer reiterated the need for students transporters and states to ask vendors for new safety technology to be added school buses via specification lists, as well as develop an understanding of how federal mandates and requirements factor in.
One concern audience members brought up was how school bus drivers are being affected by auditory input moving to visual input. For example, the familiar school bus “roar” is being replaced by much lower decibels emanating from quieter alternative-fuel school buses. A related problem is that pedestrians, especially children and the elderly, might not hear the bus operating nearby. Andersky agreed that this is a challenge that requires a solution, adding that audible alarms on school buses are prime examples of the need for the system to correctly identify the most important aspects of vehicle operation to alert the driver about, like a person too near the school bus.
The panelists all stressed that technology cannot make up for a school bus driver. It can and should only assist them. “Overwhelming someone with huge amounts of data doesn’t necessarily make them a better driver,” Alexis cautioned, advising that good interface between drivers and the data system is needed, as well as good integration between the human and automatic functions. “Don’t just throw more screens at your driver.”
Andersky agreed. “Train your drivers; don’t just throw them into a bus full of smart tech,” he concluded.