First off, Happy Veterans Day. I, like most Americans, always like to take time out every once in a while to thank all the men and women of the Armed Forces for their service. My ex-girlfriend is a Navy officer, and from what I’ve seen from her experiences, they do an absolutely astounding amount of work and truly dedicate themselves to the cause of protecting the very freedoms most Americans take for granted. Thank you.
I read an article (blog post) today from Technology Review, an online publication from MIT, about automated cars, specifically a system that would feature automated cars networked together in “road trains.” I’ve found myself driving down the highway before thinking “Wouldn’t life be better if we cleared up these highways a bit?” This sort of system absolutely fascinates me.
Ideally, you’d want to have the entire system networked together. You’d need sensors along roadways throughout the system. As the article mentions, this is tremendously expensive. The alternative is to have a professional driver in the front in a marked car. If another driver wished to follow along with them as part of a “road train”, to reduce wind drag and optimize performance and speed, they’d simply fall in behind them, or head to the back of the line, and become part of the network with that professional driver. They could then eat, sleep, read, or whatever, until they need to exit the highway, at which point they’d simply leave the “train”.
I see many engineering problems with this alternative that might make this system either unsafe or under-utilized:
For one, how many professional drivers would you have per highway? How far apart would they need to be spaced? Would this be a function of current traffic conditions? How long can a “train” effectively be?
Second, making this an optional system would mean you would still have a lot of people who don’t use it. That could present quite a hazard. Let’s say that person is unfamiliar with how a “train” works, and either cuts off the lead driver, or someone in the middle of the train? You could solve this by creating one or two designed “auto train” lanes, perhaps. Still, the danger is there.
Third, the notion of allowing drivers in the train to become distracted could be unsafe. You’d really have to be careful how you configure manual overrides and how one would be allowed to safely exit the train. Perhaps they’d need to punch in an exit number or something once they get in the train, and they’d get some sort of advisory when their exit is coming up. If they don’t respond (let’s say they fall asleep, which I’m sure can and would happen in many cases), they stay in the train.
Other considerations:
- What if a car in the train begins to run out of gas? Does the system alert the driver? What if they fell asleep?
- What if you have a network failure? Do you have a built-in failsafe strategy within each car so they perform a given action to prevent a dangerous situation?
I’m very intrigued by the idea, but if you’re going to automate, the first thing to take into account is the human factor.
I read an article (blog post) today from Technology Review, an online publication from MIT, about automated cars, specifically a system that would feature automated cars networked together in “road trains.” I’ve found myself driving down the highway before thinking “Wouldn’t life be better if we cleared up these highways a bit?” This sort of system absolutely fascinates me.
Ideally, you’d want to have the entire system networked together. You’d need sensors along roadways throughout the system. As the article mentions, this is tremendously expensive. The alternative is to have a professional driver in the front in a marked car. If another driver wished to follow along with them as part of a “road train”, to reduce wind drag and optimize performance and speed, they’d simply fall in behind them, or head to the back of the line, and become part of the network with that professional driver. They could then eat, sleep, read, or whatever, until they need to exit the highway, at which point they’d simply leave the “train”.
I see many engineering problems with this alternative that might make this system either unsafe or under-utilized:
For one, how many professional drivers would you have per highway? How far apart would they need to be spaced? Would this be a function of current traffic conditions? How long can a “train” effectively be?
Second, making this an optional system would mean you would still have a lot of people who don’t use it. That could present quite a hazard. Let’s say that person is unfamiliar with how a “train” works, and either cuts off the lead driver, or someone in the middle of the train? You could solve this by creating one or two designed “auto train” lanes, perhaps. Still, the danger is there.
Third, the notion of allowing drivers in the train to become distracted could be unsafe. You’d really have to be careful how you configure manual overrides and how one would be allowed to safely exit the train. Perhaps they’d need to punch in an exit number or something once they get in the train, and they’d get some sort of advisory when their exit is coming up. If they don’t respond (let’s say they fall asleep, which I’m sure can and would happen in many cases), they stay in the train.
Other considerations:
- What if a car in the train begins to run out of gas? Does the system alert the driver? What if they fell asleep?
- What if you have a network failure? Do you have a built-in failsafe strategy within each car so they perform a given action to prevent a dangerous situation?
I’m very intrigued by the idea, but if you’re going to automate, the first thing to take into account is the human factor.
No comments:
Post a Comment