Bosch Advances Autonomy (Even for Lawns & Parking Garages)
When you talk with Dr. Kay Stepper, head of the regional business unit Driver Assistance and Automated Driving, and vice president of engineering for the Chassis Systems Control Div., Robert Bosch LLC (bosch.us), about Bosch’s efforts in the development of automated driving, he talks about something that is completely unexpected.
Yes, there are “robotics” and “artificial intelligence.” (He notes that when he started—he joined Bosch in 1996 and his doctorate is in electrical engineering from the Technical University of Berlin, so he’s got both experience and deep knowledge—the latter was referred to as “machine learning,” a term, he suggests, dates him, although one could counter that in this fast-developing field, experience is not in the least bit a bad thing.)
Anyway, what Stepper talks about—after citing such facts that Bosch rolled out with antilock brakes in 1978, electronic stability control in 1995, 77-GHz radar systems in 2000 . . . –is a lawn mower. Yes, a lawn mower.
But this isn’t your run-of-the-mill Toro.
Rather, it is a Bosch lawn mower. A robotic lawn mower.
Back in 2014 Bosch introduced the Indego 1200 Connect, a lithium-ion-battery powered lawn mower that can be controlled by a smartphone app (iOS or Android). You think being able to honk your horn and blink your lights via an app is impressive? What about cutting your lawn?
Stepper admits that the notion of a lawn mower doesn’t sound all that exciting in the context of, say, a Jeep Cherokee that’s out in the parking lot fitted with radar, cameras, electronic power steering, electronic stability control, a motor control unit and a human-machine interface that provides lane guidance and lateral and longitudinal control so that it is able to change lanes (after driver confirmation), as well as having the ability to automatically resume from a standstill. Yet it shows a point about intelligence in motion.
The Bosch robotic lawn mower (they’ve now moved on to the Indego 400 Connect), he says, isn’t like one of those robotic vacuum cleaners that goes bumping along. “It does complete planning. It will plan a strategic pattern, then it will execute that pattern,” he says.
This is a long way from traveling on I-75 at speed. But it goes to the point of developing automated technology for the benefit of people, something that Stepper and his team, as well as many other Bosch associates, as well as partners in academia, are deeply involved in doing.
And just as you’d imagine that a Ph.D. in electrical engineering would focus on the technical challenges that are being faced vis-à-vis developing autonomous vehicle technology, Stepper says, “Yes, there are technical challenges. But I am absolutely convinced we’re going to solve them. We already have many technical elements in place.” A big challenge: “to really get the consumer to understand what the technology is and how it works is a crucial piece.”
One of the consequences of the consumer in the equation is, Stepper notes, the importance of having psychologists on his team to help develop the human-machine interfaces necessary, especially in SAE Level 3 automation, where there is “the expectation that the human driver will respond appropriately to a request to intervene.”
In Level 4, the driver may get a “request to intervene” from the system, but may not respond, so the system has to be able to handle “all aspects of the dynamic driving task.” Which leads to something that Stepper says is crucial for autonomous vehicles, which is “fail operational.” (This is also key for Level 3 automation.)
Stepper explains, “If something fails, the whole system needs to be operational.” In your car right now, the design is “fail-safe,” inasmuch as the driver is expected to step in should there be a system failure. “If your steering boost fails,” Stepper says, “you no longer have the steering amplification. But the driver is the backup: You have to provide more input, more force on the wheel, but you can still control the steering.” This is not necessarily the case when the vehicle is automated.
“So the technical solution is redundancy,” Stepper says. “We need to have redundant systems in the vehicle for sensing, steering and braking.”
Which, of course, leads to the question of cost: doesn’t all of this redundancy mean that there is the possibility of something like 2X the cost of setting up an autonomous vehicle?
No. “The backup system doesn’t have to have the same level of performance as the primary,” he answers.
This is one of the things that makes a company like Bosch beneficial for autonomous systems. That is, it is possible to use the electronic stability control system functionality to steer a vehicle by braking individual wheels. And as for braking, they have the “iBooster,” a vacuum-independent electromechanical system that can be used for backup.
“You don’t have to have two of everything.”
Perhaps somewhat more understandable than lawn mowers but still not what one might ordinarily think about when “autonomous driving” is the subject, Stepper says, “No one I have met is excited about parking. No one says, ‘I got this really hot vehicle so I can look for parking for an hour and then squeeze into a spot.' I have yet to talk to someone who says they enjoy that.”
While there are certainly benefits to be gained in terms of safety from autonomous systems (“Several studies have shown that more than 90 percent of crashes are caused by human error,” Stepper points out), there are things like autonomous parking that can make a difference: “There is not enough talk about how much inconvenience we can erase by automating that piece.”
Daimler and Bosch have developed an automated parking system that they’re piloting in a multi-story parking deck at the Mercedes-Benz Museum in Stuttgart; it will go live in 2018.
The system makes use of both on-board and infrastructure-based technologies. In operation, the driver will go to a designated area in the garage, exit the vehicle and activate the parking system. There are sensors in the garage that provide information on where there are available parking spaces. So the vehicle drives itself to the assigned parking location. Should there be unexpected pedestrians or other vehicles suddenly in the way, the automated car is capable of steering and braking as required. When the driver is ready to leave, the vehicle is recalled via a smartphone app.
Not only is this a matter of convenience, but it is calculated that the system allows 20 percent more vehicles to fit into the existing parking structure (there isn’t a need for the space between vehicles for human ingress and egress).
Bosch, incidentally, not only develops the on-board technology, but the parking garage infrastructure and communications technology that will make this first-ever (as in first anywhere in the world) automated public parking system a reality.
There have been more than 20 reported attacks against Waymo’s self-driving fleet in Chandler, Ariz., since the company began testing the technology on public roads there two years ago.
Visteon Corp. is developing DriveCore, an open platform to control and operate autonomous vehicles.
While there is a burgeoning proliferation of companies that are in the LiDAR space, each with its own take on utilizing laser pulses to create a precise map of its surroundings for purposes of ADAS or full-blown automation, a Seattle-based company has a distinction that certainly sets it apart from its competitors.