4

I recently searching Airbus Careers section, and I can say that I saw too many Cyber Security job positions. It was saying in the job description that there is a need of having knowledge about avionics and of other related sections. It was not focusing on Computer Science part.

My question is this: Is there a way that someone can hack an airplane? I mean each airplane is moving. How can someone "touch" a moving airplane, have the time to identify it and then to hack it?

Also, the airports are having really specialized machines and systems in order to communicate with an airplane. How can someone, without having specialized machines, can hack an airplane?

Machavity
  • 5,993
  • 2
  • 28
  • 57
avionerman
  • 275
  • 1
  • 4
  • 18

4 Answers4

11

The answer to

Is it possible to hack an airplane while it's moving?

is "theoretically yes, movement has no effect on a plane hackability".

How can someone "touch" a moving airplane, have the time to identify it and then to hack it?

All the people inside the airplane are touching it during the whole flight, they know which airplane they are flying with, and they have a flight worth of time.

How can someone, without having specialized machines, can hack an airplane?

Theoretically, for some components, you don't need specialized machines (see below), you need a laptop and an ethernet cable

see here (TL;DR version ), or here, and finally, from here

First, the airplanes. The problem the GAO identifies is one computer security experts have talked about for years. Newer planes such as the Boeing 787 Dreamliner and the Airbus A350 and A380 have a single network that is used both by pilots to fly the plane and passengers for their Wi-Fi connections. The risk is that a hacker sitting in the back of the plane, or even one on the ground, could use the Wi-Fi connection to hack into the avionics and then remotely fly the plane.

The report doesn't explain how someone could do this, and there are currently no known vulnerabilities that a hacker could exploit. But all systems are vulnerable--we simply don't have the engineering expertise to design and build perfectly secure computers and networks--so of course we believe this kind of attack is theoretically possible.

As we read here, though, safety critical components are not that easy to compromise:

it’s essential to understand that aircraft systems can’t be updated in-flight or on the fly. This means that if an attacker is seated on the airplane, there’s no way they could modify systems. There are only three ways to update a system on a modern aircraft:
1. Take the hardware off the aircraft and return it to the manufacturer for a factory performed update.
2. Some larger airlines have their own maintenance organizations with the required specialized equipment to perform a bench update of a particular system using a portable data loader (PDL) or some other type of specialized data-loader.
3. On-wing update [...] Specialized hardware (physical switch(es)) must be set, to place the IMA and relevant systems in maintenance mode. Once the aircraft systems are in maintenance mode, a maintenance engineer needs to follow a sequence of steps to push the update to the target system. The concept of an automatic update does not exist for any safety-critical system.

So, to summarize: yes, researchers have identified non-critical systems that could be hacked (as the passenger entertainment devices), but so far nobody has demonstrated the feasibility of hacking a critical system

Federico
  • 32,559
  • 17
  • 136
  • 184
  • 1
    Wow. I had no idea. A socking part of an article that you gave me is this: "While we appreciate responsible engagement from independent cybersecurity researchers, we’re disappointed in IOActive’s irresponsible presentation." Amazing! So yes there is a high possibility to hack an airplane. The thing is that for now there are 0 attempts. Thanks for your answer. It was more than clear :) – avionerman Aug 28 '19 at 06:06
  • 2
    I doubt the planes avionics can be affected as there is typically either an air gap or data verification/validation between systems of differing criticality. – selectstriker2 Aug 28 '19 at 13:41
  • 2
    @selectstriker2 - you'd expect that, but it is not always the case. And where there is data verification/validation, it's always possible for the systems to get it badly wrong :-) – Rory Alsop Aug 28 '19 at 15:01
  • "Remember the GAO's nightmare scenario: A hacker on the ground exploits a vulnerability in the airplane's Wi-Fi system to gain access to the airplane's network." Can a user on the Gnd even get a WiFi connection to something moving over a hundred miles an hour just before landing without being right under the plane (shortest distance)? That's a pretty big stretch. "A general rule of thumb in home networking says that Wi-Fi routers operating on the traditional 2.4 GHz band reach up to 150 feet (46 m) indoors and 300 feet (92 m) outdoors. " Planes are >300 feet up until well down the runway. – CrossRoads Aug 28 '19 at 17:36
  • 1
    @CrossRoads: The risk is more like: A passenger uses the plane's onboard Wi-Fi to browse to a rogue web site, which then causes their browser (connected to the plane's Wi-Fi, remember) to do something malicious. – Greg Hewgill Aug 28 '19 at 20:41
  • 2
    @avionerman, the IOActive's presentation is irresponsible in that it is scaremongering. They admit they don't have the engineering experience to make proper analysis of how dangerous it actually is, and they are missing many steps to get to anything actually dangerous—because the system they got in is not safety critical, so all the serious protections only start behind it—yet present it as imminent danger. But defending against this kind of PR is part of a reason why avionics manufacturers need security experts. – Jan Hudec Sep 02 '19 at 20:58
  • Your last quote seems to contradict your first point. – Notts90 Jan 10 '20 at 09:21
  • 1
    Even if a hacker can only hack the entertainment devices, the potential damage it can make in conjunction with psychological tactical to create panic in hundreds of passenger is not negligible – jean Jan 10 '20 at 14:35
4

While it may be possible for an attacker to get into certain parts of a network - for example, the infotainment system or a crew system, the avionics portion of an aircraft's system is segregated or air-gapped (depending on the aircraft's age / specification) - so, while it may be possible to interact with certain domains of the aircraft's systems there will be difficulty in transitioning from a lower-security domain such as infotainment/crew to a high-security e.g. avionics domain. This was not covered in Ruben Santamara's talk as he didn't have access to the details regarding segregation and the specifications of other parts of the network, he also wasn't working with the most up-to-date version of Boeing's code. - Further, the in-flight wifi system, while vulnerable (weaknesses were highlighted in certain systems whereby routers were discoverable on the Internet and installed with default creds), are not connected to other systems within the aircraft save for power, and as yet (as far as I'm aware) no National Aviation Authority has sanctioned the use of over-the-air updates for avionics or any other aircraft system, and all changes will still be subject to updates with safety cases etc.

Of course, this depends on the manufacturer and air carrier correctly implementing security controls.

It's also worth noting that Airbus' cyber security consultancy does not just focus on aviation, but has many contracts within defence and government.

Edit: Worth noting that larger a/c will be almost completely different in terms of avionics compared to GA which uses the CANBus system also found on road vehicles etc - however, again, physical access would still be needed.

fite
  • 41
  • 3
2

Let's distinguish the two types of "hacking" we're talking about

Total control of the airplane

This is the stuff of Hollywood and is often over-dramatized. The new season of Dr. Who started off with a car being "hacked" and trying to drive itself off a parking deck. If you're not shaking your head, you should be. This wasn't a self-driving car, it was just a regular one (it had a driver before it vaporized him with a laser from the console screen). The idea that most modern cars could be hacked to drive themselves somewhere dangerous is laughable at present. That, and hacking is so wildly overblown that it's noteworthy when someone tries to make a movie or show with realistic intrusions.

Most aircraft have two computers (at least). For simplicity, we'll call the one that controls the aircraft movement the flight control computer. Most airplanes have a way to disengage any other on-board computer and fall back to this device, which takes inputs from the control systems (i.e. the steering yoke). This computer is almost certainly not hackable because they generally don't take any input from the network. If you were going to hack this computer, you would need physical access to the plane, intimate working knowledge of the on-board computers, and a new OS/firmware you could upload into the computers (plus some way to add a way for you to control it remotely).

Messing with your head

A far more realistic scenario is that a hacker gains control of some other subsystem that is hooked to the network and wreaks havoc

In his 2014 research, Santamarta found that an in-flight airline WiFi network was vulnerable to malicious behavior via vulnerable Cobham AVIATOR 700 satellite terminals on the WiFi. The danger there was an attacker gaining control over the Satellite Data Unit or the SwiftBroadband Unit interface by taking advantage of the weak password reset feature, hardcoded credentials, or the insecure protocols in the terminal.

"More specifically, a successful attack could compromise control of the satellite link channel used by the Future Air Navigation System (FANS), Controller Pilot Data Link Communications (CPDLC) or Aircraft Communications Addressing and Reporting System (ACARS)," he wrote in his 2014 research paper. "A malfunction of these subsystems could pose a safety threat for the entire aircraft."

Imagine that ACARS suddenly starts feeding erratic data. Ground stations begin to get notices the engines have shut down or are malfunctioning. The air crew might divert out of an abundance of caution. This is costly for everyone involved, because now you're taking a plane out of service for inspection and having to reroute passengers. Worse, you have a crew potentially no longer trusting on-board systems.

These "soft" intrusions are no less dangerous, but they are very real. Often times, network security is a low priority and anything that touches the network is potentially vulnerable. You can't directly crash a plane, but messing with a crews' head can be just as dangerous.

Community
  • 1
Machavity
  • 5,993
  • 2
  • 28
  • 57
1

Yes

At Defcon this year there was an entire Aviation Hacking Village - demonstrating and discussing hack techniques and defences. The village is specifically created to identify vulnerabilities and inform manufacturers of them before they are used in the wild, allowing them to be remediated. See specifically, Harshad's talk at 1030 on Sunday for ILS system attacks, and the various CANbus exploits.

Speaking Schedule:

Friday 9th August – Bally’s Hotel Event Center – Track 1

Time Talk + Speaker

  • 1300 – 1315 Can the CAN bus fly ­Risks of CAN bus networks within avionics systems Patrick Kiley
  • 1315 – 1400 Behind the scenes of hacking airplanes Zoltan and Ben
  • 1400 – 1430 Hacking the Air Force and Beyond Dr. Will Roper and Jack Cable
  • 1430 – 1500 A Hacker Walks Into A Flight School And Says Ouch: Common Online Security Fails In Pilot Training Tarah

Saturday 10th August – Bally’s Hotel Event Center – Track 1

Time Talk + Speaker

  • 1000 – 1100 Panel – The Long Haul: The State of Aviation Security Policy Andrea, Stefan, Pete, Renderman
  • 1100 – 1200 A hackers first solo: airplane avionics security 101 Ken and Alex

Sunday 11th August – Bally’s Hotel Event Center – Track 1

Time Talk + Speaker

  • 1000 – 1030 Ideas whose time has come: CVD, SBOM, and SOTA Katie and Art
  • 1030 – 1100 Wireless Attacks on Aircraft Instrument Landing System Harshad
  • 1100 – 1130 In The Air And On The Air: Aviation Radio Systems Exploding Lemur
  • 1130 – 1200 An introduction to the ARINC standards Karl

Demos and Show-And-Tell:

Date / Time Event

  • All Day / Every Day CANBus Avionics Demo Patrick from Rapid 7
  • All Day / Every Day PTP Workshop
  • All Day / Every Day Bricks in the Loop DDS
  • All Day / Every Day ADS-B Workshop Renderman and Jim Workshop
  • Sat 10th / 1100 Talk: A Venture into the Dark Side John McCarthy Workshop
Rory Alsop
  • 1,475
  • 16
  • 29
  • "Fail" is a verb not a noun. – quiet flyer Aug 28 '19 at 13:27
  • 3
    @quietflyer, "Fail" can certainly be used as a noun these days – selectstriker2 Aug 28 '19 at 13:38
  • 3
    Without making a judgement on whether commercial aircraft are vulnerable to hacking, a convention on a topic does not make that topic plausible. For example, see any conspiracy theory conventions. Therefore, I wonder how does this answer the question? – AEhere supports Monica Aug 28 '19 at 14:54
  • 2
    It's not a convention on the topic - that was part of the world's biggest hacking convention, demonstrating attacks, so it absolutely answers the question. Or maybe I should just start with the word- Yes :-) – Rory Alsop Aug 28 '19 at 14:58
  • 2
    I haven't heard of anyone demonstrating anything, except in simulators – selectstriker2 Aug 28 '19 at 15:09
  • 1
    Well - it's really worth following the links on that page to their twitter, and I think videos. I think there were two valid vulnerabilities over the last couple of years (It's not my field, I'm more critical infrastructure and private sector) – Rory Alsop Aug 28 '19 at 15:14
  • Answers should stand on their own, if something was demonstrated at this convention, you should mention it. – AEhere supports Monica Aug 28 '19 at 16:10
  • @AEhere - I'm happy to take it to chat, either here or over on security.stackexchange. Please read the list of talk titles above, and be aware that there have been valid (although limited attacks) by ethical hackers already, as noted by Federico. The Defcon ones are new - and thankfully discussed with the manufacturers so they can fix before publication. – Rory Alsop Aug 29 '19 at 09:26
  • 2
    @RoryAlsop I don't think I can add much beyond the points I already have raised: this is a list of panels about hacking, that does not imply they come to positive or practical conclusions, if they did have a notable result to show, it should be explicitly linked. As it stands this is a link-only answer where you have quoted the less useful part of the linked content. Contrast with Federico whose answer links directly to the claims and counterclaims, and quotes a specific vulnerability concern. – AEhere supports Monica Aug 29 '19 at 09:39
  • @AEhere - I disagree, the titles are the key items. As I stated, the talks are about the attacks and defences developed. They do have practical conclusions, which help the industry improve, and I'd suggest it is worth your attention. But you are entitled to your opinion, so have a good day. – Rory Alsop Aug 29 '19 at 09:45
  • 2
    I have to agree. This is not an answer. I have read a few of the reports from Defcon this year and they did indeed find some pretty important potential vulnerabilities. But this doesn't detail any of them. Its just a schedule of what various people spoke about, and an invition to go look in another castle. – Jamiec Aug 29 '19 at 09:53
  • 1
    @RoryAlsop I, in turn, disagree that "An introduction to the ARINC standards" or "Hacking the Air Force and Beyond" give any meaningful insight into the kind of conclusions one can expect from those panels. Even I could host something that fits under the first one, given some time to brush up, and not mention hacking once. This topic does have my attention, hence my interest in this question; my gripe here is with the way in which your information is presented, which is vague even by aviation hacker claim standards. If they do have practical conclusions, why not link or summarize them? – AEhere supports Monica Aug 29 '19 at 09:53
  • Because the presentations are not published online yet – Rory Alsop Aug 29 '19 at 11:40