Learning from aviation - PART 2 - human factors
This is the second part of the blog discussing what lessons learned in aviation industry can be applied to space industry. In aviation we have a saying that "all the rules have been written in blood", meaning that many lives have been lost in the process of evolving the barnstormers into passenger jets where old grannys drink cofee at 12000 m altitude and speeds of 900 km&h and more.
This part of the blog is dedicated to human factors.
"To error is human" is the basic assumption of aviation industry.
Many decades ago people realized that human beings are not perfect and that they are more or less susceptible to errors and imperfections.
Here are some major errors and human imperfections commonly known as "human factors" with description and mitigation solutions:
1. physical and pscychomotorical imperfections.
Humans are not capable to perform any physical or psychomotorical activity with 100% perfection. For example, when flying without autopilot, the pilots cannot keep the assigned altitude, speed and course "to the feet, knot and degree". To try to make pilots perfect flying machines is a contraproductive "macho" approach, often being the cause of accidents or incidents. The solution is to provide tolerances for each phase of the flight. Therefore, if we allow the altitude to be maintained within a certain range (e.g. +/- 100 ft) we can realistically expect that pilots will be within this safety zone, provided adequate training is being applied. We can then use this "buffer zone" as the worst case scenario when designing rules for separation between airplanes and terrain etc.
This way we have designed our procedures taking into account the possibility of a human error instead of expecting the humans not to make errors.
Similar good examples are adding buffer on go/no-go decision speed during takeoff, design of navigation procedures taking into account "overshoots" during turns, errors in timing, speed, course and altitude etc.
Space industry should take into account all such human imperfctions and design the systems around the limitations that we have. This will prevent situations in which we are unable to perform because we are simply, physically limited.
There is a specialised section of aviation dealing with fatigue management, duty time and rest limitations. Human performance is significantly reduced in case of fatigue and fatigue cannot be avoided at all times. A good example of human factors management is to create a procedure for inflight rest of pilots. The pilots can get tired at any phase of the flight, and by not allowing them to rest, we get a contraproductive situation. They are going to rest anywy, but in an uncontrolled way. The better option is to acknowledge that at certain times pilots need to "close their eyes" for 20 minutes and make strict procedures how to do it in a safe way, for example, stating the duration limitations, requiring that the other pilot takes control over all aircraft operations and to notify the flight attendant that the procedure is in force.
2. mental imperfections
Human mind does wonderful things, but the mental processes have so many imperfections that we don't even recognize most of them.
We forget things, we miscalculate things, we remember the things in a wrong way, we omit things, we get distracted, we have negative transfer of habits from previous airplanes/cars/machines, we fight with our egos, don't admit mistakes etc.
How do we make a flight safe if we have 5 volumes of airplane systems description, procedures, limitations, 4 volumes of company rules and regulations, 20 volumes of international regulations, and dozens of local regulations on one side, and our aviation worker, the average Joe on the other side?
Firstly a good aviatior can discriminate between the "need to know" and the "nice to know". The most important capacity of an airman is to discriminate important things from unimportant.
Secondly, there are some nice tools assisting us to overcome the problems which may occur due to human factors.
One of the basic tool is the check list. Simply put, a check list is a list of actions relevant to a certain portion of a mission, but there is much more to it than meets the eye. Special rules are written how check lists are used. For example, if more than one person is involved in an activity, such as a passenger jet, each item in the check list has a defined rule who reads it, who responds and how the correct action is confirmed. Each checklist has a rule who calls it, who reads it and who responds. Most check lists have two ways of performing them - "Read and do" and "Do and read", meaning that they can be either performed by reading and doing each item step by step, or all the items can be done by heart and then read and checked by the check list.
Here is an example of an MD-80 normal check list. You can notice that each item has a correct reponse and a person who does it (CM1/CM2 - left seated/right seated pilot, PF/PNF - pilot flying/pilot monitoring)
Normal check list prevents us from forgeting itmes and allows us to perform the actions in the correct order.
Some of the biggest accidents happened due to poor normal procedures, or due to not complying with them.
In former USSR, testing of a multistage rocket went horribly wrong because normal procedures were skipped on site and batteries and ignition systems were being tested at the same time. This caused all stages of the rocket to ignite at the same time causing several hundreds of deaths.
Another great example is damaging the oxygen tanks on US Lunar missions because the heaters were tested with the wrong power supply.
As the system develops, the experts must develop the normal procedures as well.
Non normal procedures are designed to anticipate any possible abnormality or emergency situation. If such a situation can be anticipated, then enough time and resources can be invested into finding the best possible solution. This is very helpful in case an emergency occurs and the operators or pilots have to make a quick decision which can have long term consequences on the mission and / or safety. Usually, the non normal procedures have a "memory items" part, followed by a "read and do" check lists and ending with the consequences and options. Memory items are performed by memory in case an immediate action is required, such as extinguishing a fire. Furhter procedures are done in order to secure the system and "save what can be saved". At the end we have the consequences and options, such as list of lost equipment, list of available systems, and various mitigation options.
Here is an example of an engine fire procedure (memory items marked with a star):
Additional things taking into account the human factors are designing the navigation charts in a user friendly way, drawing attention to important information and discretelly providing non critical information etc.