b'85should the vehicle remain? And is it okay to pass that cy-clist, or is it better to slow down and even stop and wait? The more complex the scenario, the more variables there are to weigh. The SDS brings all these variables together to choose an action that is natural, something that both passengers and other road users will understand. Once the system has sensed its surroundings, predict-PLANNING ed future behavior, and planned its optimal trajectory, its In a perfect world,everything on the road wouldtime to act.follow a set of rules and predictable behaviors. The real world, however, is never this simple. In order to plan how it will get from Point A to B, the Argo SDS has to answer two sets of questions as it drives down the road: What is everyone else going to do next? And how can it use thisACTING information to plan a safe, natural motion towards itsIn order to act,the Argo SDS is equipped with destination?motion-control capabilities that take action on the tra-To get started, the Argo SDS generates a series of pre- jectory. They direct the engine, braking, and steering sys-dictions about the future movement of the objects it istems as the vehicle moves comfortably and safely along a tracking. For instance, while a car driving straight willplanned path. often continue driving straight, the SDS will also predictThemotion-controlprocessconsiderstheforce-tens or even hundreds of possible futures, like that carbased dynamics of moving a car, and how they factor into making a right turn at an upcoming intersection, or merg- driving naturalistically. To move (or stop) a thing, you ing into the lane to its left. Because it is aware of the pos- have to apply a force. The SDS considers all the real phys-sibility of a right turn for the other car, the SDS can adjustics that impact how a self-driving vehicle starts, stops, or its speed or lane positioning in anticipation. This is an- changes direction on its journey from A to B. For exam-other way that Argo delivers naturalistic drivingsubtleple, cars traveling at 35 mph cannot instantaneously come actions that help maintain a smooth experience.to a complete stop: they need time and distance for the While predicting the potential paths of other objectsbrakes to slow the car to a stop. The SDS factors every-is important, the Argo SDS goes further: it predicts be- thing, including the amount of brake needed and the time havior as well. These ever-changing predictions are basedit takes to come to a complete stop.on a variety of inferences about the objects surroundingFinally,becauseeverycityisunique,anddriving the vehicle. Predictions consider where an object is fac- norms and laws vary from a place like Miami to Austin ing and the direction in which its moving (a car can beto Pittsburgh, Argos SDS behavior can be tuned to local facing one direction but moving in reverse). They alsodriving culture. For instance, Pittsburghhome of Argo take into account scene-level context. For instance, theAIs headquartersis known for the famous Pittsburgh SDS can detect a crane, a truck, and traffic cones, and in- left, during which oncoming traffic gives up the right-of-fer that its in a construction zone. This in turn aff ectsway and politely lets left-turning vehicles turn against a how the SDS estimates the behavior of other actors in thegreen. The Argo SDS is trained to anticipate local behav-scene. The SDS can also infer that a person crossing theior like the Pittsburgh left in order to yield if appropriate.road far away from a crosswalk may be trying to get to aIn considering any journey of a self-driving vehicle, car parked on the other side. naturalistic driving and safe driving go hand-in-hand. After the Argo SDS takes in its surroundings, it plansGradually inching (instead of darting) into traffic before its own trajectory for the next few seconds. As it plans,making a turn, or easing into a new lane on the highway the SDS is searching for the best action to take next,instead of swerving rapidlythese are the subtle, but considering what it is already doing (driving down theessential,distinctionsthatmakeself-drivingvehicles road, say, or waiting at a red light), all the predictionsbehavenaturalistically.Ifitspleasingforridersand of what other road users are going to do, and how theyrecognizableforotherroadusers,theneveryoneis might interact with the self-driving vehicle. made safer.As the Argo SDS plans its path to its next destination, it needs to maintain safety, comply with the law, and drive naturalistically. If the vehicle approaches a stop sign or red light, for example, it requires itself to stop. It will not change lanes if there is another actor next to it. And while waiting at a stop sign to cross a road with the right of way, the SDS will either wait for traffic to clear, or recog-nize that it must clear the intersection to avoid forcing a crossing vehicle to slow down unnecessarily. These are all things the SDS must identify and consider in its motion.The rest of the planning process involves evaluating answers to a slew of additional questions. Is it better to slow down, to shift positions in a lane, or to actually change lanes in response? How far away from a cyclist FullBook_Mar24.indb 85 4/25/21 6:43 PM'