Why Self-Driving Cars Must Be Programmed to Kill in the Most Ethical Manner

Self-driving cars are already cruising the streets. But before they can become widespread, carmakers must solve an impossible ethical dilemma of algorithmic morality.

As reported by MIT Technology Review: When it comes to automotive technology, self-driving cars are all the rage. Standard features on many ordinary cars include intelligent cruise control, parallel parking programs, and even automatic overtaking—features that allow you to sit back, albeit a little uneasily, and let a computer do the driving.

So it’ll come as no surprise that many car manufacturers are beginning to think about cars that take the driving out of your hands altogether (see  “Drivers Push Tesla’s Autopilot Beyond Its Abilities”). These cars will be safer, cleaner, and more fuel-efficient than their manual counterparts. And yet they can never be perfectly safe.

And that raises some difficult issues. How should the car be programmed to act in the event of an unavoidable accident? Should it minimize the loss of life, even if it means sacrificing the occupants, or should it protect the occupants at all costs? Should it choose between these extremes at random?

The answers to these ethical questions are important because they could have a big impact on the way self-driving cars are accepted in society.  Who would buy a car programmed to sacrifice the owner?

So can science help? Today, we get an answer of sorts thanks to the work of Jean-Francois Bonnefon at the Toulouse School of Economics in France and a couple of pals. These guys say that even though there is no right or wrong answer to these questions, public opinion will play a strong role in how, or even whether, self-driving cars become widely accepted.

So they set out to discover the public’s opinion using the new science of experimental ethics. This involves posing ethical dilemmas to a large number of people to see how they respond. And the results make for interesting, if somewhat predictable, reading. “Our results provide but a first foray into the thorny issues raised by moral algorithms for autonomous vehicles,” they say.

Here is the nature of the dilemma. Imagine that in the not-too-distant future, you own a self-driving car. One day, while you are driving along, an unfortunate set of events causes the car to head toward a crowd of 10 people crossing the road. It cannot stop in time but it can avoid killing 10 people by steering into a wall. However, this collision would kill you, the owner and occupant. What should it do?

One way to approach this kind of problem is to act in a way that minimizes the loss of life. By this way of thinking, killing one person is better than killing 10.

But that approach may have other consequences. If fewer people buy self-driving cars because they are programmed to sacrifice their owners, then more people are likely to die because ordinary cars are involved in so many more accidents. The result is a Catch-22 situation.

Bonnefon and co are seeking to find a way through this ethical dilemma by gauging public opinion. Their idea is that the public is much more likely to go along with a scenario that aligns with their own views.

So these guys posed these kinds of ethical dilemmas to several hundred workers on Amazon’s Mechanical Turk to find out what they thought. The participants were given scenarios in which one or more pedestrians could be saved if a car were to swerve into a barrier, killing its occupant or a pedestrian.

At the same time, the researchers varied some of the details such as the actual number of pedestrians that could be saved, whether the driver or an on-board computer made the decision to swerve and whether the participants were asked to imagine themselves as the occupant or an anonymous person.

The results are interesting, if predictable. In general, people are comfortable with the idea that self-driving vehicles should be programmed to minimize the death toll.

This utilitarian approach is certainly laudable but the participants were willing to go only so far. “[Participants] were not as confident that autonomous vehicles would be programmed that way in reality—and for a good reason: they actually wished others to cruise in utilitarian autonomous vehicles, more than they wanted to buy utilitarian autonomous vehicles themselves,” conclude Bonnefon and co.

And therein lies the paradox. People are in favor of cars that sacrifice the occupant to save other lives—as long they don’t have to drive one themselves.

Bonnefon and co are quick to point out that their work represents the first few steps into what is likely to be a fiendishly complex moral maize. Other issues that will need to be factored into future thinking are the nature of uncertainty and the assignment of blame.

Bonnefon and co say these issues raise many important questions: “Is it acceptable for an autonomous vehicle to avoid a motorcycle by swerving into a wall, considering that the probability of survival is greater for the passenger of the car, than for the rider of the motorcycle? Should different decisions be made when children are on board, since they both have a longer time ahead of them than adults, and had less agency in being in the car in the first place? If a manufacturer offers different versions of its moral algorithm, and a buyer knowingly chose one of them, is the buyer to blame for the harmful consequences of the algorithm’s decisions?”

These problems cannot be ignored, say the team: “As we are about to endow millions of vehicles with autonomy, taking algorithmic morality seriously has never been more urgent.”

Ref: arxiv.org/abs/1510.03346 : Autonomous Vehicles Need Experimental Ethics: Are We Ready for Utilitarian Cars? 

Transcontinental US Record Aboard Tesla Model S P85D with Autopilot Claimed

As reported by GizMag: Earlier this year, Carl Reese and Deena Mastracci made a 3,011-mile (4,846-km) drive from Los Angeles to New York in 58 hours and 51 minutes in a Tesla Model S P85D to set a record for the coast-to-coast journey for an electric vehicle. Now Reese and Mastracci, joined by Alex Roy, have beaten that record with the help of Tesla's new Autopilot software.

According to the Transcontinental Drivers Association, the yet-to-be-verified record-breaking drive left Redondo Beach, California on October 18 at 9:15 pm PST and arrived at the Red Ball Garage in New York on October 21 at 10:03 am EST – making for a distance of 2,995 mi (4,820 km) and a time of 57 hours and 48 minutes.

The drive relied on the newly-released Tesla Autopilot software, which was made available by software update to Tesla S and X owners on October 15. This provides semi-autonomous capabilities so the vehicle can be driven without touching the pedals or steering wheel on the highway.

The car uses cameras and radar to keep an eye on other vehicles, potential hazards and road markings. It can execute lane changes with only driver supervision, dodge swerving vehicles, park automatically, and prevent drivers from making potentially dangerous moves like driving over the divider.

The Association says that both the departure and arrival of the Tesla S were recorded by witnesses and the GPS tracking data will be submitted to verify the record.

"Deena and I have been planning this autonomous record for over 11 months," says Reese. "We knew we wanted to be the first group to make the cross-country attempt using the autonomous technology."

London's Iconic Black Cab Goes All Electric

As reported by GizMag: The London Taxi Company (LTC) has unveiled a new version of the Black Cab that famously operates in London, UK. The TX5 is said to retain the spirit of past models and focus on driver and passenger comfort. Reflecting the move toward a low-carbon economy, it will be electric powered.

The predecessor of the new London taxi, the TX4, was launched in 2007 and, although electric versions of it were developed, models on the streets invariably had a conventional diesel engine under the bonnet.

According to LTC, all of the 2,000 components for the TX5 have been designed from the ground up. It has a new lightweight aluminum body structure and composite paneling, rear-hinged doors to improve passenger access and space for a forward-facing wheelchair. It also has a panoramic glass roof.

The cab can accommodate six passengers and provides more legroom for the driver. It is equipped with Wi-Fi connectivity and mobile device charging points for both the driver and passengers.

The TX5 was designed by Brits Peter Horbury and David Ancona at Geely’s design center in Barcelona. They were supported by over 200 engineers and designers in the UK. Hornbury says they were acutely aware of the responsibility that came with designing a new model of such a recognizable and iconic vehicle.

"We wanted to retain those aspects that make the London taxi what it is and what it stands for, while ensuring that the new design reflects everything that a 21st century vehicle needs to be," said Horbury. "We have sought to build on existing features such as the headlights, grille and strong roof lines. But we have also made sure that the new vehicle significantly builds on space and comfort levels for both passengers and drivers."

More detailed specifications about the TX5 have yet to be announced. It is due to be launched in the UK at the end of 2017 and across international markets in 2018.

Battelle’s DroneDefender Anti-Drone Beam Gun Grounds UAVs

As reported by GizMag: There's recently been a run of new anti-drone systems introduced to deal with potential threats from UAVs, but these have been on the large and expensive side. To provide an affordable alternatives to plug the gap between shotguns and truck-mounted systems, national security research and development firm Battelle is introducing DroneDefender. Billed as the first portable, accurate, rapid-to-use UAV counter-weapon, it's a rifle-like raygun device that uses a radio beam to jam drone control systems and stop them in midair.

Most UAVs in service today are relatively harmless, but their growing numbers increases the risk that they will be used maliciously or carelessly. Recent systems for dealing with rogue drones have taken one of two approaches. One type destroys UAVs by shooting them down with special anti-aircraft guns, which has the obvious limitation that such weapons can't be used in many areas for legal or safety reasons. The other type uses large-scale jamming systems that fire radio beams to interfere with the drone's controls, but many of these, like the anti-aircraft guns, are very large, expensive, and not at all suitable for dealing with minor problems at small installations or following VIPs and other targets on the move.

Battelle's DroneDefender is designed as a point-and-shoot system that looks like an elaborate shoulder raygun with two antennae, a software-defined radio, and jamming circuitry. It works by firing a radio beam in a 30° cone that jams the control and GPS navigation frequencies to disable drones at distances of up to 400 m (1,300 ft). Battlle says that the system not only freezes the UAV, but also stops all outside control commands, including radio detonation signals.

The DroneDefender places a heavy emphasis on speed and portability. It weighs under 10 lb (4.5 kg), can cold-start in under a tenth of a second, and operates continuously for up to five hours with optional battery packs. It can be carried or fixed on a Picatinny rail mount and comes in hardened versions for forward deployment in hazardous areas.

Battelle says that the DroneDefender requires no special training to use, has already been tested against a DJI Phantom drone target, and has shown a consistent response in field conditions.

Because, US FAA regulations prohibited an actual demonstration, the video below shows a simulation of the DroneDefender in use.

Bulldozers Guided by Drones are Helping Ease Japan's Labor Shortage

As reported by The Verge: Komatsu is the world’s second largest construction company, a venerable Japanese brand with 94 years of history that sells forklifts and bulldozers to customers around the globe. But in its home country, Komatsu has been struggling with an aging population, a trend that has left few young workers available to operate its machines.

As Japan ramps up new construction in preparation for hosting the 2020 Olympics, experts believe it will face a serious obstacle. "The labor shortage in the construction industry could reach a crisis level in the next few years," Martin Schulz, an economist at Fujitsu Research Institute in Tokyo, told Bloomberg.

To get around this problem, Komatsu has begun creating a new service it calls Smart Construction. A team of robotic vehicles scoops rock and pushes dirt without a human behind the wheel. They are guided in their work by a fleet of drones, which map the area in three dimensions and update the data in real time to track how the massive volumes of soil and cement are moving around the site.The drones are built by an American startup, Skycatch, and today it revealed that Komatsu is one of the principal investors in a new $25 million round of funding that the company plans on using for global expansion and R&D. The partnership will power an entirely new line of business for Komatsu: Instead of selling construction equipment to its customers, it can lease them out for a job, and use a staff of remote operators and autonomous vehicles to complete the work. One or two lightly trained humans are still required on site to take control if something goes wrong.

Before switching to drones, Komatsu had been experimenting with autonomous dump trucks, bulldozers, and excavators, but they lacked the ability to see and understand the environment around them with enough precision to be useful on their own. Komatsu would use teams of human surveyors to create extremely detailed maps of the job site, a process that left a lot of room for improvement. "Because the terrain survey could not be conducted with high accuracy, what results in many sites is a 20 to 30 percent margin of error in soil volume after the construction is complete," says Chikashi Shike, an executive with Komatsu’s Smart Construction division.With Skycatch drones, Komatsu says it has dramatically reduced that margin of error while dramatically cutting the time it takes to complete a sitemap. "With the former, traditional method, it takes about two weeks, on average, to survey a certain piece of land," says Kenishi Nishihara, a project manager with the Smart Construction division. "Meanwhile with Skycatch it can be completely down within one day, or even 30 minutes."

Skycatch is based in Silicon Valley, but has decided to focus its business almost entirely on Japan. "The regulations in the United States make it difficult to operate in a fully autonomous manner, and so that makes it tough to get the full value out of drones," says Skycatch CEO Christian Sanz. Of course, not every country is facing such dramatic labor shortages. "In Japan, there is a clear recognition, and appreciation, of the work these drones can do," says Sanz. "In other places, people are still cautious about collaborating with robots to do jobs that were typically handled by humans."

Unmanned System for Navy Boats Takes to the Water

As reported by GizMag: In the waters near Portsmouth Naval Base, a small black boat recently roared about with no one at the controls. It hadn't run amok after the pilot fell overboard, but was instead a demonstration of a new robotic system developed by ASV and BAE Systems. The technology package can be retrofitted to the Rigid Inflatable Boats (RIB) used by the Royal Navy (RN) to turn them into high-speed, autonomous, unmanned reconnaissance and surveillance platforms.

Even in an age of aircraft carriers and satellites, small boats are still vital Naval assets. The manned Pacific 24 RIB used by the RN aboard its Type 23 Frigates and Type 45 Destroyers will also be used by the Queen Elizabeth Class aircraft carriers now under construction. They act as the eyes and ears of their parent ship by going where she cannot. Unfortunately, it's often dangerous work and the Royal Navy and other services are rather keen to find ways to send in the boats while the sailors stay home.

This is where the new autonomous demonstrator comes in. It's designed to not only to take the sailor out of the equation, but also to be retrofitted to the Pacific 24s already in service.

The unmanned system and its software were developed by Portchester, UK-based ASV Unmanned Marine Systems with BAE Systems helping to integrate the technology in the demonstrator. BAE Systems says the robotic boat can reach speeds of over 38 knots (44 mph, 70 km/h), and operate for up to 12 hours and up to 40 km (25 mi) from its ship. It can be controlled by either a conventional human pilot or a hand-held remote controller, or run completely autonomously.

For navigation and reconnaissance, the system uses a sensor suite that includes navigation radar, a 360-degree panoramic infrared camera array, and a laser range finder. These are backed up by complex algorithms developed by ASV.

"The algorithms we’re developing with BAE Systems allow the boat to perform complex missions and navigate through waters avoiding collisions," says Dan Hook, Managing Director for ASV. "This gives it the flexibility and sophistication to operate in a number of different tactical roles, whether it’s patrolling areas of interest, providing surveillance and reconnaissance ahead of manned missions, or protecting larger ships in the fleet."

According to BAE Systems, the demonstration opens the way to the next phase of development, which will see the creation of the sensor suite and it's integration into the parent ship combat management system.

"This technology delivers an extremely robust and fast-moving unmanned boat that is able to perform a number of surveillance and reconnaissance roles, even when operating at high speed or in choppy water," says Les Gregory, Product and Training Services Director at BAE Systems. "BAE Systems has a wealth of experience in the development and integration of unmanned systems. The successful demonstration highlights the enhanced capability this technology offers. While other programs are primarily designed for larger, slower boats to tackle mine counter-measure scenarios, this system provides an extremely maneuverable multi-role vessel."

The video below introduces the new autonomous system.

Stanford Turns a DeLorean Into a Drifting, Driverless Car

As reported by Engadget: This isn't from some weird movie franchise mashup entitled Back to the Future IV: Stanford Drift. No, Stanford's Revs Center simply decided to experiment on an old DeLorean, giving it the ability to fly drift all on its own. The car, which the team decided to called MARTY (Multiple Actuator Research Test bed for Yaw control) as an homage to B2F, was heavily modified and modernized. DeLoreans are old cars, and they found that it had severe limitations: for instance, it was understeering, so they had to equip it with a power steering motor and rack.

As for why they wanted to design a driverless car that's good at drifting in the first place, well, Revs Center director Chris Gerdes told Wired: "We think automated vehicles should be able to execute any maneuver within the physical limits of the vehicle to get out of harm's way." In other words, they believe autonomous cars should be able to drift if it means saving the passenger's life. If the idea of a drifting DeLorean sounds weird, just watch the team do some perfectly executed donuts in the video below.