TECH TALK
issues, but this is also handling the resulting chaos better and minimizing the eff ect upon passengers and cargo. In fact, passengers are appreciative if airlines are able to make arrangements ahead of time such as contacting them to inform them of impending delays so travel plans can be proactively managed.
OPERATING AIRCRAFT Some form of automated systems have been used in commercial aviation for many years, and have arguably made fl ying safer than ever. Automated fl ight systems have practically become a co-pilot (minus the good banter). Newer systems such as Maneuvering Characteristics Augmentation System (MCAS) have emerged to help provide greater safety when there’s a need for adjustments due to aircraft handling characteristics. Such systems use various sensor data to automate the management of control surfaces of an aircraft due to fl ight conditions. This was not wholly possible prior to better sensors being added/upgraded to aircraft, and for larger data buses to move more data to fl ight control systems, and for AI algorithms to be integrated into cockpit systems. Aircraft such as the Airbus A350 twin-jet have 50,000 sensors which collect ~2.5 terabytes on a typical fl ight day, and newer aircraft from Boeing must have similar capabilities. Since MCAS is a newer system,
there may have been issues, and the catastrophic Lion Air 610 illustrates this. According to published reports, the pilots were unaware of the capabilities of the anti-stall system, which were largely omitted from the plane’s operations manual (the investigation is still underway). The Lion Air Boeing 737 MAX 8 that crashed shortly after taking off from Jakarta, Indonesia on Oct. 29, 2018, had activated an automatic trim
24
DOMmagazine.com | apr 2019
system based on erroneous angle of attack (AOA) information. The pilots were without an optional feature that would have alerted them to the error, according to a preliminary report issued by the Indonesian National Transport Safety Committee. Situations such as this illustrate that not only the need to properly document new systems is sorely needed but more training as well. Even trained pilots cannot be expected to adjust to entirely new systems without proper support, especially when lives are at stake. A proper balance needs to be
struck between assisting fl ight crew with safely operating aircraft and becoming dependent upon such automated systems. And this is true for any AI-driven system in any industry or home. When sensors fail, and systems need to be overridden, users need to be completely aware of the possibilities of such situations. In order to better design new AI systems, more stakeholders need to be involved as early as possible to create solutions which accommodate as many viewpoints as possible. Airbus has taken a novel approach by establishing the Airbus AIGym. This is an online platform that focuses on Airbus-specifi c challenges (the 3rd challenge was recently fi nalized) to be solved with artifi cial intelligence (AI) and machine learning (ML). This latest challenge is to identify new and unexpected changes in the behavior of monitored systems, as well as analyze suspicious behavior for potential faults and failures more effi ciently and more quickly. Airbus opens this to everyone; companies, start-ups, research labs, schools or individuals. The recently started contest is utilizing three datasets covering cases in the helicopter, satellite, and commercial aircraft business domains.
In 2018, Boeing launched a
new business called the Disruptive Computing and Networks (DC&N), which is tasked with research and development of solutions in artifi cial intelligence (AI), secure communications and complex systems optimization for commercial and government applications. This new venture will interact with the company’s investment arm, HorizonX, to identify outside companies to collaborate with. It was reported that DC&N will focus on computing and neuromorphic processors for handling complex problems and pattern detection, and add other areas over time. The real-time analysis and optimization of global air traffi c routes for both manned and unmanned vehicles is one of the early projects. In late 2018, Boeing and SparkCognition announced the launch of a joint venture SkyGrid, which will build an AI- and blockchain-powered airspace management software platform. SkyGrid will create a software platform to ensure the safe, secure integration of passenger air vehicles and autonomous cargo operating in the same airspace. SkyGrid’s platform will be based on secure blockchain technology, and use AI-enabled dynamic traffi c routing and data analytics to provide greater capabilities beyond today’s unmanned aircraft systems (UAS) traffi c management (UTM). The platform will enable SkyGrid customers to safely track unmanned aerial vehicles in fl ight and allocate traffi c corridors for autonomous vehicles. The NASA Ames Research
Center, located in Silicon Valley, is working on a project on identifying “anomalous operations” using datasets from commercial aviation, which could provide early warning of serious issues. According to a recent
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68
