For safety critical airborne systems, the minimum reliability requirements are of the order of one allowable failure in one million flight hours. This implies that almost two thirds of cost and time for design and development is devoted to complying with certification requirements of the aviation regulatory authorities. The recommended design practices have two distinct categories. The system safety analysis which ensures that with reference to high level requirements, no possible failure condition has been overlooked. The physical design has to make sure that no failure condition will occur beyond specified level of probability. Thus the system level design requirements are derived from system safety analysis.
The system safety analysis is based on qualitative parameters which cannot be mathematically modeled. Also it requires in-depth knowledge of aeronautics. These processes are covered in the Society of Automotive Engineers Aerospace Recommended Practices ARP-4754(Guidelines For Development Of Civil Aircraft and Systems) and ARP-4761(Guidelines and Methods for Conducting the Safety Assessment Process on Civil Airborne Systems and Equipment). At present only large aircraft manufacturing companies have some proprietary software for handling safety critical airborne equipment. The focus of this startup is to automate these processes using Artificial Intelligence and Machine Learning techniques, so that companies without sufficient knowledge of aeronautics can venture into development of safety critical aerospace systems.
The overall system comprises a set of software that can accept as input the textual high level requirements and then go through the complete process of system safety analysis. This will comprise generating Functional Hazard Assessment(FHA) from the stated requirements. From the generated FHA the software carries out Fault Tree Analysis(FTA) or Markov Analysis(MA) to perform Preliminary System Safety Analysis(PSSA). Derived from high level requirements calculate system utilization as per phases of flight and the same is used to determine and probability of failures are thus calculated. The software can carry out Failure Mode and Effects Analysis (FMEA) to determine the effect of any component failure on the system. The independence of systems and functions from failures in associated systems can be carried out using graphical analysis to fulfill the requirements of Common Cause Analysis (CCA). Thus, by taking the high level system requirements as input the complete system safety analysis will be performed and supporting artifacts will be generated. This will greatly cut down on development costs and time.
The aviation market is projected to grow many times in the coming decade. Furthermore, the concept of air taxis and Urban Air Mobility will increase the number of aircraft being produced exponentially. All these aircraft will require certification of safety critical systems. Consequently, the number of users requiring such services will keep growing.
Aviabot is a startup launched to provide these services and training to potential developers, online and onsite, so that they can reduce their time to market their products and be competitive in the global aviation market. The startup has seasoned aeronautical engineers and young avionics engineering students who are working to develop these algorithms and software.