Recovering Aircraft Safety after Loss of Pilot Control: An Innovative Algorithm

N Ananthkrishnan, Department of Aerospace Engineering


Over the recent years we have witnessed frequent media reports on crashes of military aircrafts in our country. Such incidents do keep recurring globally in peacetime. Crashes could be due to various causes: bird hits, mechanical defects, bad weather, etc. However, recent statistics have shown that a large number of crashes are due to a specific problem faced by pilots called spatial disorientation (SD).

When flying difficult sorties and under poor weather conditions, pilots can be confused (disoriented) about which way they are heading (up or down), and whether the ground is below their feet or above their head! For example, military pilots are known to suffer from visual illusions during night flying such as mistaking discrete ground lights for the stars and consequently flying inverted (upside down).

A recent study has shown that almost 90-100% of aircrew have reported at least one incidence of SD during their flying career. Pilots either fail to recognize an SD condition and hence take no corrective action or, even when they recognize the problem, are too disoriented to be able to recover the aircraft to safe flight. In most cases, the aircraft ends up in what is called a spin or a spiral dive with the pilot having no control of the aircraft – the airplane nose drops, it starts going around in circles while losing height rapidly.

Spatial disorientation is a problem that can confront any pilot, no matter how highly experienced and well trained. During the years 1980-89, the US Navy reported 112 major accidents, and the US Air Force reported 270 major accidents, involving SD and loss of pilot control. Pilots of general aviation (light) aircraft are equally vulnerable to SD - one of the more high profile crashes was that of the Piper Saratoga being flown by John F Kennedy, Jr. on July 16, 1999. Unfortunately, many accidents caused by spatial disorientation are wrongly labeled as due to pilot error.

To avoid loss of costly airplanes and to save precious human lives, a two-pronged strategy has been suggested in the literature:

  • Pilots should be trained in flight simulators to recognize SD situations and hit a Panic Button provided in the cock pit
  • The aircraft's automatic flight control system should have a Panic Button Algorithm that takes control of the air plane from the pilot and recovers the airplane to a nor- mal flying condition.

However, developing an effective Panic Button Algorithm has been a challenge because of the tight constraints involved: pilots will usually hit the button only when they are in a hopeless situation with the plane already hurtling to the ground, and the algorithm must respond in a very short time before an imminent crash.

The New Algorithm
In a major breakthrough, researchers at the Department of Aerospace Engineering, IIT Bombay, working over the last 3 years (2002-04), have come up with a novel Panic Button Algorithm that seems to meet the challenges pointed out above. The research team consisted of students (P K Raghavendra and Tuhin Sahai, P Ashwani Kumar), a research assistant (Manan Chauhan), and the author. The work was partly funded by the Aeronautical Development Agency (ADA), Bangalore.

Using a combination of two sophisticated new methods called Nonlinear Dynamic Inversion (NDI) and Extended Bifurcation Analysis (EBA), the team from IIT Bombay has devised a unique Panic Button Algorithm that successfully recovers an airplane from even the most adverse flight conditions. The crux of the present work lies in recognizing that a successful algorithm must use a two-step approach where it is necessary for the airplane to pass through an intermediate (waypoint) state before it can be properly recovered to a safe flight condition.

The research team has carried out extensive computer simulations using high-fidelity aerodata obtained from NASA for a specially modified F-18 airplane called the High Angle-of-Attack Research Vehicle HARV (see illustration) to establish the effectiveness of their algorithm. In the future, the Panic Button Algorithm could be built into sophisticated Flight Control Systems being developed for advanced combat aircraft such as 'Tejas' the Indian Light Combat Aircraft (LCA). Interestingly, their work also shows that aircraft equipped with thrust vectoring (TV) engines, such as the Sukhoi SU-30, have a 60 per cent better chance at successful recovery as compared to aircraft without TV capability. Translated in terms of height from the ground, airplanes with TV can be recovered after loss of control at much lower altitudes, which is important since nearly 100 per cent of loss of control cases at low altitudes presently end up as crashes.

Presented at the Aerospace Sciences Meeting organized by the American Institute of Aeronautics and Astronautics (AIAA) at Reno, NV, USA (Jan 2004), the work has been appreciated internationally for its thoroughness and novelty. It is expected to be of high value to the international aircraft design community.