the good word groundswell

Friday, August 18, 2017

Russia May Be Testing Its GPS Spoofing Capabilities Around The Black Sea

A consensus is building that Russia has been experimenting with the ability to override Global Positioning System signals with fake ones that can severely disrupt and confuse the geolocation abilities of GPS users. For now this phenomenon has mainly affected ships traversing the Black Sea, but during a time of conflict it could impact a wide spectrum of combat systems including drones, missiles, smart bombs, as well as surface combatants and land warfare units. Even manned aircraft may not be immune to such electronic warfare tactics.

This all started on June 22nd when a ship near the Russian port of Novorossiysk reported a major miscalculation with their high-end GPS installation. New Scientist states:

"On 22 June, the US Maritime Administration filed a seemingly bland incident report. The master of a ship off the Russian port of Novorossiysk had discovered his GPS put him in the wrong spot – more than 32 kilometers inland, at Gelendzhik Airport.

After checking the navigation equipment was working properly, the captain contacted other nearby ships. Their AIS traces – signals from the automatic identification system used to track vessels – placed them all at the same airport. At least 20 ships were affected.

While the incident is not yet confirmed, experts think this is the first documented use of GPS misdirection – a spoofing attack that has long been warned of but never been seen in the wild."

Maritime-executive.com has a more thorough account of the incident you can read here. The commander of the ship that reported the strange occurrence documented it thoroughly.

Russia's large port at Novorossiysk sits at the northern end of the tense Black Sea region, near the Kerch Strait and across from the Russian-held Crimea. Novorossiysk is a major port for (the) Russia Black Sea Fleet.

As New Scientist points out, GPS spoofing seems to be something Russia has been playing with in a major way before this anomaly occurred in the northern Black Sea region. Reports that GPS users around Moscow, and especially near the Kremlin, have had similar issues, with the peculiar GPS behavior being discovered in a most unusual way:

"Over the past year, GPS spoofing has been causing chaos for the receivers on phone apps in central Moscow to misbehave. The scale of the problem did not become apparent until people began trying to play Pokemon Go. The fake signal, which seems to center on the Kremlin, relocates anyone nearby to Vnukovo Airport, 32 km away. This is probably for defensive reasons; many NATO guided bombs, missiles and drones rely on GPS navigation, and successful spoofing would make it impossible for them to hit their targets."

Insidegnss.com elaborates further on Russia's affinity for GPS denial tactics:

"With more than 250,000 cell towers in Russia equipped with GPS jamming devices as a defense against attack by U.S. missiles, the country has advanced capabilities to disrupt GPS. There have been press reports of Russian GPS jamming in both Moscow and the Ukraine. In fact Russia has boasted that its capabilities “make aircraft carriers useless,” and the U.S. Director of National Intelligence recently issued a report that stated that Russia and others were focusing on improving their capability to jam U.S. satellite systems, according to the RNT Foundation."

GPS jamming, or the loss of GPS altogether, has long been a major concern of the US military. One highly publicized yet at the same time shadowy drill led to warnings that GPS blackouts and disruptions would emanate from the Mojave Desert and reach hundreds of miles in every direction. Other elaborate testing and systems development programs have increased in frequency and scope regarding operating in GPS denied environments. This includes a cocktail of new technologies that can offer similar geolocation capabilities to GPS without the need for a constellation of satellites above.

Still, wide-area GPS jamming or outages are easy to detect, but spoofing GPS signals is a much more insidious threat, especially if spoofing effects are localized in nature. New Scientist states:

"...It can now be done with commercial hardware and software downloaded from the Internet. Nor does it require much power. Satellite signals are very weak – about 20 watts from 20,000 miles away – so a one-watt transmitter on a hilltop, plane or drone is enough to spoof everything out to the horizon."

Advanced guidance and navigation systems, like those found on aircraft and in most GPS-guided weapons, use inertial navigation with embedded GPS. System software loaded onto the INS/GPS guidance system uses algorithms to detect discrepancies among various steams of information. This way the system can place less priority on one stream of information, say from a malfunctioning ring-laser gyro or GPS receiver, than others. Or it can discount that steam entirely if it starts straying far from the consensus of the others. This way a navigation system that suddenly loses a component due to failure or has one of its data streams deviate for other reasons won't simply stop working.

For instance, if the GPS telemetry suddenly leaps miles away, the navigation system may "vote out" the GPS information entirely. The system will be less accurate overall because of it, but it will still function. As an example, take a Joint Direct Attack Munition, the most common guided bomb used by the USAF, which uses GPS and INS navigation. If the system suddenly loses GPS connectivity during its attack sequence, INS will take over, but the bomb will be substantially less accurate than it would have been with both GPS and INS working together as a team.

JDAM has a tailkit with moving fins, a GPS antenna and a GPS/INS autopilot.

But spoofing isn't as simple as telling a GPS receiver it is somewhere totally different than where it really is. By broadcasting false GPS data that slowly changes over time, navigational systems with GPS and INS may not "throw out" the GPS data so soon, or at all for that matter. Over time this could result in vehicles going far off course or even running into shores, mountains or even other vehicles without warning. Some say this is how the Iranians brought down the RQ-170 Sentinel drone in a relatively intact state in late November 2011, although this remains highly debatable.

As for what emitter or platform that may have caused the disruption, we don't know for sure. But it is interesting to note that a Tu-214SR strategic communications relay aircraft was deployed to Crimea shortly before the incident took place, and has been flying regular circles over the northern Black Sea ever since. It's possible the aircraft is capable of rebroadcasting fake GPS signals aside from its primary duties, or at least is experimenting with the capability in region.

As you can see, this new form of electronic warfare is powerful and potentially quite deadly. The big question is how or when would Russia use it as a weapon? It seems that it fits well into Moscow's hybrid warfare playbook, as it could disrupt the navigation of unfriendly forces and even impact their ability to target Russian forces. It could also disrupt commerce and air travel all while the hostile emitter stays safely in Russian territory.

On a peer state warfare level, such as in a conflict against NATO, this technology could not only protect Russian forces from kinetic attacks by weapons that use GPS, but it could also disrupt any allied operations that relies heavily on GPS for navigation today. Even logistics to support such an operation could also be negatively affected.

Basically, the US military, and much of the world's transportation and commerce, runs on GPS. If anything else this incident in the Black Sea is a stark reminder of why the Pentagon needs to wean itself off this dependency as soon as possible. This possible event also doesn't take into account other avenues of attack a foreign power has to disrupt the GPS network, which includes cyber-intrusion or even attacking GPS satellites directly. Anti-satellite capabilities have become far more exotic, hard to detect, and more numerous than they once were. And even electronic warfare targeting GPS satellites doesn't have to be terrestrial in origin. Jamming tactics can be deployed in space between orbiting objects.

Among all this uncertainty, one thing is for sure—during a major conflict, GPS will be a prime strategic target. Considering the technology to disrupt or spoof its signals is becoming ever more accessible, non-state actors will likely take advantage of it too for nefarious purposes. For instance, just think of this tactic's implications when it comes to a concept like self-driving cars.

With all this in mind, preparing for a world where GPS connectivity is not guaranteed may not be convenient, but it may be entirely necessary.

Wednesday, August 16, 2017

Tesla's Upcoming Electric Semi Truck Will Be Able to Drive Itself

Autopilot, big rig style.

As reported by The Verge: Tesla CEO Elon Musk has been teasing an electric semi truck for a while now, ahead of an official unveiling this fall. But a report in Reuters adds a new, if somewhat unsurprising, wrinkle to the mix: the Tesla big rig is probably going to have self-driving capabilities.

Reuters has seen emails between Tesla and the Nevada DMV where the two sides discussed “potential road tests” of the truck’s self-driving capabilities. The information also apparently describes Tesla’s desire to create long-haul electric semis that can drive themselves in “platoons,” potentially following behind a lead truck piloted by a human driver.

The idea that Tesla is working on incorporating self-driving technology into its upcoming semi truck falls in line with how aggressive the company has been at building the same tech into its consumer cars. Tesla offers semi-autonomous features on all of its current models in the form of Autopilot, which costs an additional $5,000 at the time of purchase. It also offers a $3,000 “full self-driving” option, which the company says will be activated once the software is ready. (Tesla claims that all of its cars are already equipped with the hardware necessary for full autonomy.)

What is surprising is that Tesla appears ready to test this technology. The state of Nevada is a likely partner, as it’s one of the few in the country that actually gives out licenses for autonomous vehicle testing. It was also the first state to allow self-driving big rigs to test in 2015 when Daimler acquired two AV licenses for its own Freightliner Inspiration Truck. Volvo is working on adding autonomous capabilities to its own trucking fleet, too.

Autonomy is also a common theme for the Silicon Valley companies that have dipped into the trucking world. In 2016, Uber acquired self-driving truck company Otto, which was led by a former high-profile employee of Google’s own self-driving project. (That employee is now currently at the center of a legal battle between Google’s parent company and Uber.) And Waymo, the company that blossomed out of Google’s self-driving car project, is working on its own self-driving truck program.

Tesla is planning an official reveal of the semi truck in September, so that’s when we’ll likely learn just how far the company wants to push this new part of its self-driving ambitions.

Monday, August 7, 2017

You Can Confuse Self-Driving Cars by Altering Street Signs

As reported by Engadget: While car makers and regulators are mostly worried about the possibility of self-driving car hacks, University of Washington researchers are concerned about a more practical threat: defacing street signs. They've learned that it's relatively easy to throw off an autonomous vehicle's image recognition system by strategically using stickers to alter street signs. If attackers know how a car classifies the objects it sees (such as target photos of signs), they can generate stickers that can trick the car into believing a sign really means something else. For instance, the "love/hate" graphics above made a computer vision algorithm believe a stop sign was really a speed limit notice.

It's easy to see the potential problems. You could make these stickers using a printer at home, so anyone from dedicated attackers to pranksters could try this. It might lead to a crash the moment someone alters the sign, but it could also produce long-term chaos -- picture your city closing a road until maintenance crews can scrape the stickers off a sign.

There are ways to fight this. The research team suggests using contextual information to verify that a sign is accurate. Why would you have a stop sign on the highway, or a high speed limit on a back road? We'd add that local governments could also install signs that use an anti-stick material, or put them out of reach. Whatever happens, something will have to change if passengers are going to trust self-driving cars' sign-reading abilities.

Of course another alternative would be to teach the AI systems to recognize sarcasm when they see it.

Thursday, August 3, 2017

Lockheed to Build $350M Satellite Factory Near Denver

As reported by the Gazette: Aerospace giant Lockheed Martin Corp. said Wednesday it plans to spend $350 million building a production facility for satellites at its Waterton Canyon campus in the Denver area that is designed to boost production capacity, reduce costs and accommodate a growing workforce.

Construction is scheduled to start by the end of the month and be completed in 2020 on a 266,000 square-foot building named Gateway Center that will be one of the largest production facilities on the campus. The facility will be large enough for the company to build five of its flagship A2100 satellites at the same time.

The new project is also designed to reassert the company in a sector that has become a lot more competitive, so-called microsatellites, which are often just a few meters wide.

U.S. Air Force officials talked in April at the Space Symposium in Colorado Springs about smaller, Swiss-Army-Knife-style microsatellites with multiple capabilities that can be quickly rocketed to the heavens to replace bigger satellites damaged in war.

Space "is a big place, but we plan to be a leader in it," said Rick Ambrose, executive vice president for Lockheed Martin's space systems business. "We're trying to develop a capability to handle any direction the market moves."

Plans for the new facility call for manufacturing, assembling and testing centers that are large enough to keep multiple production lines running at the same time, allowing the facility to produce large satellites as it is churning out smaller ones.

Lockheed executives are hoping that that will allow them to compete in the small-satellite market at the same time that they keep legacy production lines going, all under one roof.

By putting assembly and testing under the same roof, employees can simply roll a satellite down the hall versus moving it between buildings for testing. By taking the steps out of packing the satellite, moving it and unpacking it, the company will cut testing time from about two days to under an hour.

The building also will include a thermal vacuum chamber, meant to simulate the harsh environment of space, that will be used to test fully assembled satellites before they are attached to rockets. The facility is to include a stock of 3-D printers to produce the satellites' simpler components.

The construction project will create 1,500 jobs over three years. Lockheed Martin currently has about 8,600 employees in Colorado, including 4,000 at the Waterton Canyon complex. The company has added more than 750 jobs to its Colorado workforce since 2014 and has 350 job openings in Denver. In April, the Colorado Economic Development Commission approved $12.6 million in tax credits for the company in return for bringing 550 new jobs to Jefferson County over eight years.

The project will bolster Colorado's reputation as a leader in the space industry. Colorado has the nation's second-largest aerospace economy, employing 188,280 people and more than 400 companies, with a $3.4 billion annual payroll, according to the Colorado Space Coalition. Gov. John Hickenlooper said at the April Space Symposium that the state has ambitions to take over the top space economy ranking.

Besides Lockheed, seven other top aerospace contractors have significant operations in the state, including Boeing and Raytheon. Colorado is also home to the Air Force Space Command in Colorado Springs and the University of Colorado's Laboratory for Atmospheric and Space Physics in Boulder.

Lockheed has played a role in space operations since the 1950s, when its rockets were used to launch some of the first American satellites into orbit. Satellites are still a critical part of the company's business, and the project is designed to reassert the company in a sector that has become a lot more competitive.

Lockheed's business is being challenged by a cadre of well-heeled newcomers that have pledged to populate the lowest rung of Earth's orbit with microsatellites.

The global telecommunications investment firm SoftBank is spending $1.7 billion to combine Luxembourg-based Intelsat with Richard Branson-backed OneWeb, creating a satellite behemoth that seems singularly focused on smaller, cheaper models. Elon Musk's SpaceX is working to launch 4,425 small satellites into low Earth orbit by 2024.

These competitors contend that their microsatellites can provide a better signal in more places.

Then there is the Air Force's interest.

Air Force Space Command has been eyeing small satellites for years and more recently has taken serious steps to integrate them into future war plans.

While less capable than the school-bus sized spacecraft the military now has in orbit, the diminutive birds are cheaper to build and faster to launch than their complicated cousins.

With nations including Russia, China, Iran and North Korea developing or possessing anti-satellite capabilities, the small satellites could be used as stopgap measures to fill in for larger spacecraft damaged or destroyed in a war that reaches orbit, Space Command boss Gen. Jay Raymond has said.

It remains to be seen how Lockheed will fare. It specializes in taking on the kind of big projects that few companies have the scale to complete. The satellites Lockheed makes circle the planet much farther outside Earth's atmosphere, giving them a broader view of planet's surface. That means they have to be much more powerful, entailing longer production times and more expensive components.

Wednesday, August 2, 2017

What are the Scary Ripple Effects of Self-Driving Vehicles?

As reported by ReadWrite: By 2040, we think fifth-level, autonomy-enabled mobility will be available as a service for the majority of the transportation needs of urban consumers. In other words, 70% of the urban population wouldn’t need to own cars because they would be available on-demand through their favorite app.

Although several power players like General Motors and Ford are promising fully autonomous cars sooner, there are many technological and large-scale regulatory and consumer sentiment issues with autonomous cars that would need to be addressed before they could fulfill our transportation needs.

But imagine the world in 2040, when most of the population doesn’t need cars. Everything from shopping to commuting to long distance road trips will be addressed by fully autonomous vehicles that you can summon with the push of a button.

Rethinking today’s products and services in the 2040 world

Once you accept that basic premise, it is surprising to see how so many of our current products and services do not fit well with the world of the fully autonomous car. When we shift from complete ownership to an on-demand leasing model, automotive OEMs and their supply chains (Tier 1 and Tier 2) will be the first ones affected. Several of them are thinking of strategies to grapple with this eventual reality, but this is just the tip of the iceberg. A significant number of multibillion dollar companies operate around consumers owning at least one to two cars per family. All these products and services will either become obsolete or have to be fundamentally rethought.

15 minutes for an insurance product we don’t need?

Let us start with the way we buy cars — at dealerships. Once consumers stop buying cars, there is no need for dealership networks. They are merely a distribution channel that will be replaced by an app that we use to hail our cars.

Then, there’s insurance; we get insurance as soon as we buy a car. Because we will no longer own cars, there is no need to spend 15 minutes to save on car insurance. Accidents because of systemic failures like poor connectivity or algorithmic edge cases are inevitable, but the massive reduction in the frequency of accidents coupled with insurance being bundled into the mobility service will result in an enormous reduction in revenue for the insurance industry.

Car loans, which finance our car ownership, will also be out. In the 2040 world, mobility service providers will end up owning most of the cars. Even if you assume one or two ride hailing apps will dominate the market, there are economies of scale in owning and operating large autonomous car networks. As a result, we believe the ultimate owner of the cars will be large businesses, not individual car owners who would lease their cars on ride sharing platforms. These large fleet owners would have a low cost of capital with which to finance their vehicle purchases. Through both a massive reduction in the number of cars and the APR they can charge for each car loan, auto loan providers will see a huge downsizing of their market.

First stop on the road to oblivion: gas stations

Once we buy our cars, we spend a lot of money on it. Let us think about these products and their relevance in the 2040 world.

The first and most obvious car-related expense is gas. In a world where we have autonomous car fleets operating, we don’t need as many gas stations. A few large gas stations far away from high-density areas and perhaps operated, once again, by large fleet owners should suffice. These cars would essentially act like public transportation vehicles, which are refueled at the end of each day at a central location. A fewer number of miles traveled because we will be using ride sharing services, with more fuel efficient cars will certainly be a cause for concern for oil companies.

Parking is another common expense with a poor consumer experience. The omnipotence of autonomous ride-sharing cars will lead us to have fewer parking lots away from high-density areas where mobility service providers can park their automotive fleets when demand is low. Close to half of our urban areas are dedicated to parking, and therefore, a massive reduction in parking lots will be a boon to our cities.

Every big technological change has massive unforeseeable consequences

At conception, the Internet was a platform only for email. But its unintended consequences ended up creating companies like Amazon, Google, and Facebook, which have changed the way we buy and consume media. In the same vein, we believe autonomous mobility will fundamentally change the way we live in our citie1s. So much of our urban real estate is tied to car ownership model of today: parking lots, gas stations, dealerships.

No one can foresee the new products or services that could emerge from such surplus real estate becoming available or the effects it will have on our housing markets. Will people live much farther from cities because commuting in an autonomous car could be productive, or would urban housing become cheaper because we don’t need parking lots anymore?

And as for my personal favorite unintended consequence, short distance flights — would we rather take a flight from San Francisco to Los Angeles or would we prefer to get there in swankier-than-business-class autonomous vehicles that come with plush beds and Netflix? I am extremely confident that there are many more positive, albeit unintended, consequences of autonomous cars that will emerge en-route to 2040.

So why is this scary?

The 2040 world of autonomous mobility is scary because so many of today’s products and services would have to radically evolve to stay relevant. And it is not just the automotive OEMs that are in trouble — auto insurance companies, car loan providers, oil and gas companies, car dealerships, parking lot owners, and auto parts suppliers and stores are all on the chopping block. Just this foreseeable disruption alone is worth $2 trillion in terms of products and services we consume today. If these companies are affected, it will set off a chain reaction of problems for suppliers, which will trigger panic.

This $2 trillion will be reshuffled and distributed to consumers, new companies and incumbents. And that is scary. Some of the best incumbent players in all the industries highlighted above are already thinking ahead to prepare and adapt to new reality. But many others will likely not survive this disruption. That’s scary, too.

Preparing for the 2040 future now by partnering with startups and augmenting your organization with change makers that can imagine the future well before it has arrived is essential if incumbent companies want to survive in this changing landscape. But for the startup founders and venture capitalists involved in the industry, the evolution of autonomous cars is an enormous and exciting opportunity that has the potential to create multiple $1 billion technology companies. Uber and Lyft are just the beginning.

Tuesday, August 1, 2017

China: Keeping Track of Technology for Self-Driving Vehicles

As reported by the Shanghai Daily: The next generation of cars, already on the drawing boards, require considerable time, research and testing before they become commonplace on the roads.

To place itself at the vanguard of “smart cars,” Shanghai created a test ground known as the National Intelligent Connected Vehicle Shanghai Pilot Zone — a 100 square-kilometer site in the Jiading District that is the first of its kind in China.

The name is a bit of a mouthful, but it simply means a demonstration site where the latest innovations in smart car technology can be tested under real-life conditions.

It’s all part of China’s ambitious plan to become a world leader of a new generation of cars that operate on digital intelligence systems, including self-driving. The target to complete the plan is the end of 2025. It’s also part of the nation’s program to upgrade its prime industries to compete in a changing world.

The pilot zone in Shanghai, opened in June last year, has been a testing site for more than 10 major car makers and auto-parts companies, including SAIC, General Motors, Volvo, Ford, BMW, NIO and Delphi. The testing covers a range of technologies, including self-driving capability, sensor performance, lane adherence and high-definition map positioning.

“Companies are actively participating in the development of the pilot zone,” said Rong Wenwei, general manager of Shanghai International Automobile City, which oversees the zone. “What we are building is a platform for companies to accelerate development of the industry.”

Ford Motor Co, for one, is testing its driver-assisted technologies, including left turn assist and traffic light optimal speed advisory. The US company has said it hopes both features will figure in its next-generation vehicles.

“The Shanghai International Automobile City provides a setting where we are able to develop, test and refine future connected vehicle technologies,” said Trevor Worthington, vice president of product development at Ford Asia Pacific.

Left turn assist is a feature that uses information exchanged between vehicles to alert drivers of oncoming traffic when making a left turn.

Traffic light optimal speed advisory technology connects a vehicle to road infrastructure, informing a driver of the best speed to reduce the waiting time at stoplights by monitoring data from roadway devices. Ford said that the advisory could help drivers avoid red lights, thus reducing travel time by up to 20 percent.

“Imagine your daily commute with less waiting time for red lights,” said Thomas Lukaszewicz, manager of automated driving for Europe and China at Ford. “Vehicle-to-infrastructure technology under development holds great promise to make commuting smoother and less time consuming.”

Delphi, the global auto parts supplier, has demonstrated its self-driving technologies in the zone. The company applied nine radar units and cameras in the autonomous car.

The radar is used to capture information about the surrounding environment of a vehicle. The camera is used to capture the road conditions in front of the vehicle.

“Chinese consumers are willing to embrace new solutions especially Internet and tech-based autonomous driving, ” said David Paja, president of electronics and safety at Delphi.

Delphi said it has achieved rapid progress on advanced driver assistance technology in China in the past five years.

“We are actively in discussion with several original equipment manufacturers in local market,” said Frank Wang, president of electronics and safety business for Asia Pacific at Delphi.

In addition to the vehicle testing, the pilot zone is a useful tool in raising public awareness about what’s coming next in the driving experience. An area of the zone has been set aside for public education. It offers a 70-minute tour on what it will be like to drive intelligent, connected cars. Visitors can also take rides in self-driving cars, including the Tesla Model X, Volvo XC90 and Volvo S90.

“If I get the chance to visit the pilot zone, I want to take a ride in a self-driving car,” said Wang Xing, a student from Tongji University. “I am curious about autonomous driving technology and want to see how it works.”

The pilot zone has itself a series of goals for the next three years. By 2020, the zone plans to host more than 1,000 intelligent connected vehicles.

The zone also aims to shoulder at least five national major projects and develop more than 10 industry standards related to intelligent and connected vehicles. It is on track to recruit more than 100 top professionals and serve as an incubator to 30 innovative startups in the three-year period.

As a hub of domestic research and development, it also plans to strengthen work relationships among companies, universities and institutions.

Shanghai International Automobile City said it has signed agreements with more than 20 companies to conduct 100 research projects. It also plans to hold more than 50 professional seminars on intelligent and connected vehicles this year, focusing on testing technologies, standards and big data.

Monday, July 31, 2017

Self-Driving Car Demo is the First to Cross the US-Canada Border

As reported by Engadget: As a rule, self-driving car tests tend to be limited to the country where they started. But that's not how people drive -- what happens when your autonomous vehicle crosses the border? Continental and Magna plan to find out. They're planning to pilot two driverless vehicles all the way from southeastern Michigan to Sarnia, Ontario, making this the first cross-border test of its kind. The machines won't be in complete control for the entire route, but they'll use a combination of cameras, LiDAR and radar to take over when they can, including two key border crossings (the Detroit-Windsor Tunnel and the Blue Water Bridge).

This isn't the first autonomous driving-related agreement involving Michigan and Ontario, but it's an important one: it'll explore rules and regulations in addition to the usual self-driving data collection.

As you might guess, tests like this will be vital to making autonomy a practical reality. Driverless vehicles need to know how to adapt to changing road rules, such as different signage and units of measurement. While this isn't the greatest challenge, it has to be overcome if you're ever going to embark on cross-border shopping trips without touching your steering wheel.