Operational Safety Cases - From the Countryside Into The Cities

How does the drone industry process?

With the evolution of UAV technology and a better understanding of the risks involved in their use, the UAV market is reaching a transformative point in its journey. Industries such as inspection, surveying and entertainment have been harnessing the power of an aerial viewpoint, but the growth of these markets has largely been limited by geographical constraints. To obtain transformative change in our world, we need to start looking at the next steps in the evolution of this technology. To achieve the long term goal of operating UAVs routinely in congested areas by companies such as Amazon and their highly anticipated Prime Air service [1], the next step appears to be moving operations from the countryside into the cities. But what have we learnt so far? How can this transformation occur? How can we manage the risks when operating in built-up environments?

Author: James Dunthorne

Published: 30 April 2017

From The Countryside...

The use of UAVs has largely seen growth in controlled environments, most notably oil and gas, construction, utilities and cinematography. The main advantage with these markets is the location of the work. Oil rigs are in the middle of a vast ocean, construction sites and film sets are normally controlled sites and separated from the public, and the majority of electricity pylons and pipelines run through non-populated areas of the countryside. These markets have become fiercely competitive and we are now reaching a maturity in these sectors. This begs the question...where is the industry heading next?

Into The Cities...

UAVs tend to find most of their uses when videoing, inspecting, and modelling interesting and transformative areas of our world. Whether this is videoing a city skyline, modelling a site to enable computer aided design before a construction or refurbishment project or managing and documenting assets post construction. Most of these areas tend to be in and around congested areas, and so restricting the use of UAVs in these areas restricts their potential. Traditionally there has always been a nervousness from both operators and regulators regarding flying UAVs in congested areas, mainly due to a lack of technology maturity and understanding of the risks. Congested areas by definition are areas where there are lots of people and properties, and so a crash in one of these areas has a far higher chance of hurting someone or damaging property. So how can this be done safely?

The Operational Safety Case

Although operating in a congested area is technically permitted with a UAV with a mass of less than 7 kg, practically complying with the rules of a standard Permission for Commercial Operation (PfCO) is very prohibitive for a majority of projects. This is mainly due to having to maintain large distances from the public and from 3rd party property both during take-off and in flight. To help move UAV operations from the countryside into the cities, back in 2014 the Civil Aviation Authority (CAA) introduced a scheme called the Congested Area Operational Safety Case (CAOSC) [2], now named the Operational Safety Case (OSC). To do this operators need to submit a three volume Operations Manual [3] which includes:

  • Volume 1: Company Information

  • Volume 2: Equipment Information

  • Volume 3: Specific Operations Risk Assessment (SORA)

Volume 1 includes information about the companies structure, staff competency, training requirements and operating procedures. Volume 2 includes information about the equipment including its specification, safety systems and evidence regarding its reliability. Volume 3 is a risk assessment detailing the exemption from the rules which the operator is requesting and an associated risk assessment showing how any additional risks mitigated.

For example, if a company wanted to extend the distance at which it could fly a UAV from the pilot, Volume 3 would need to explore the risks associated with doing this, and explain how they have been mitigated. In this case, there would be an increased risk of not being able to avoid other aircraft if the UAV were on a collision course. To mitigate this risk the company could refer to information contained within Volumes 1 & 2 and include evidence to show how the ability to see and avoid other aircraft can be maintained at extended distances [4]. The OSC submission needs to use the "as low as reasonably practicable" (ALARP) methodology of reducing risk. This weighs up the risk versus the trouble, cost and time needed to control it.

The Exemptions

In order to operate routinely in congested areas, the following rules in the Air Navigation Order (ANO) are prohibitive:

1. The person in charge of a small un-manned surveillance aircraft must not fly the aircraft within 50 metres of any vessel, vehicle or structure which is not under the control of the person in charge of the aircraft.

2. During take-off or landing, a small un-manned surveillance aircraft must not be flown within 30 meters of any person [5].

Finding an area to take-off and land which is 30m from a person, prohibits routine use in congested areas. This limits operations to sites which are either segregated from the public (such as construction sites) or which are situated next to a park or similar. This is hardly going to push UAVs into the cities routinely.

Secondly data that is to be acquired generally needs to be high resolution photography, whether that is for image based modelling or inspection purposes. To obtain this, the UAV will generally need to be flown at heights lower than 50m and this means that the 50m rule will be prohibitive in a lot of projects.

Therefore it is evident that these two rules need to be tackled in any OSC. A reduction in these distances opens up a whole new stream of operating locations.

Operating Overhead People and Property

Recently the CAA stopped issuing PfCO's with the "directly overhead" wording in their clauses. Their stance is that the wording does not stand up to legal scrutiny. Many people may now think that this provides an open door for pilots to operate overhead people and property not under their control, however the CAA refers us to Article 94 (2) and Article 241 in the ANO [5]. These read as follows:

94. (2) The person in charge of a small unmanned aircraft may only fly the aircraft if reasonably satisfied that the flight can safely be made

241. A person must not recklessly or negligently cause or permit an aircraft to endanger any person or property

Both of these articles essentially put the responsibility of ensuring safety onto the operator.

Article 94 (2) appears to refer to pre-flight checks and site assessments. The pilot must have done the necessary checks to ensure that the aircraft is in a good state of health and the flight is safe to carry out given the operating location. They pilot must be satisfied that the flight can be completed safely.

Article 241 seems to refer more to the completion of a comprehensive risk assessment. This risk assessment must look at the probability of different modes of failure, the potential consequence of those failures and therefore the risk involved in the flight. If the risk is deemed too high, then mitigating factors should be put in place to ensure that performing the flight would not be deemed reckless.

As an example, operating an aircraft with just 4 motors and with no data on the reliability of those motors or their mean time to failure, means that the probability of a failure of any motor could be deemed quite high. When hovering above people, the chances of a failure leading to an impact with a person would be quite high. If the UAV had a take-off mass of 5 kg, the consequence of a failure could be considerable. Taking all of this into account a lot of operators may deem this flight too risky and so either the equipment they were using would need to be revised, or other forms of mitigation would need to be in place (i.e., a ballistic parachute) . The only way this will be tested will be in a court of law with a judge staring the operator in the eye, and so it is important that all operators do everything they can to conform to the ALARP methodology of reducing risk.

So what are the key risks when operating in congested areas?

Key Risks

Below is a list of key risks which has been compiled over many years of piloting of which a large proportion have been in and around central London. Some of these may be obvious but others less so:

  • Failure of a motor - if no redundancy is available, then this will lead to a crash with a high chance of hitting either a person, vehicle or property not under the control of the pilot

  • Power failure - losing power to the UAV will generally result in the aircraft falling vertically towards the ground with potentially devastating consequences

  • Failure of autopilot - there are many things which can fail on an autopilot and this can lead to unstable behaviour and a crash

  • Loss of line of sight - it is very easy to lose the UAV over the horizon when operating in built up areas. Generally you are surrounded by buildings, and if the pilot is not concentrating it can be easy for the UAV to be lost behind a rooftop resulting in a collision.

  • Loss of GPS - losing GPS is never a good thing and will result in the pilot having to take over manually to rectify the situation. If the pilot is not trained well or not concentrating this can result in a crash. Flying in and around tall buildings increases the chance of either losing GPS, or even worse getting GPS reflections which can cause unstable behaviour.

  • RF interference - cities are full of Wifi signals and so it is far more common to experience loss of links. A loss of link should result in the aircraft returning to home, however this can be problematic especially when facing GPS issues. If a tall building is between the UAV and the home point, this return to home sequence can result in a collision.

  • Magnetic interference - metal in the ground as well as sources of high voltage DC power can cause unstable flight characteristics. Due to these sources being near the ground, they present a risk during take-off and landing of a collision with a nearby object or person.

  • Wind turbulence - when operating close to tall buildings and in high winds, UAVs can be subject to pretty fierce wind turbulence from air being forced around buildings. This can cause the aircraft to momentarily become uncontrollable resulting in a crash.

  • Birds - a whole range of different birds live in our cities, some of which like to flock in large numbers and some of which are rather territorial. The risk of a collision with a bird is something which many people do not consider, but is actually quite significant.

  • Collision with buildings - operating in congested areas with tall buildings is difficult at the best of times. A momentary lack of concentration can cause a pilot to misjudge distances between the UAV and buildings and result in a collision.

  • Collision with aircraft - operating in cities such as central London, there is a greater volume of air traffic compared to other places. Low flying helicopters are fairly common, and if flying at high altitudes the risk of a collision is much greater than in more rural settings.

  • Distraction of drivers - when operating close to busy roads, it is easy for drivers to become distracted especially if they see a UAV hovering by the side of the road.

There are many other risks that could be included here, but this covers the most common ones. The next thing to consider is how we can mitigate these risks...

Platforms, Personnel & Procedures

There are three aspects within a business which enable companies to control risks. The equipment that is used, the staff that are employed and the procedures that are followed. Using poor quality or badly maintained equipment can lead to malfunction. Poorly trained or unfit staff will be less able to operate equipment safely. Badly thought out procedures will lead to errors being made.

This philosophy of risk mitigation will allow a company to reduce risk to acceptable levels. Here are a few key things to consider for each of the 3 P's:


  • Redundancy of systems

  • Safety features

  • Reliability of systems

  • Maintenance (initial, continued and continuing)

  • Inspections such as pre-flight checks

  • Testing of batteries and firmware


  • Training (internal and external)

  • Qualifications (external)

  • Experience (hours)

  • Currency (platform dependant)

  • Health (general and eyesight)

  • Condition (working hours, drugs and alcohol)


  • Risk assessments

  • Checklists (pre-flight, unpacking, packing away)

  • Exclusions of operations

  • Site assessments

  • Communications (police, ATC)

  • Limitations (weather, platform)

  • Logs and records (battery voltages and resistances, maintenance records, flight logs)

By considering all of these factors, risk can be reduced to something that would be deemed manageable.


It has been shown that UAVs are increasingly starting to be used in cities, and this trend is only likely to gain momentum. Through the mechanism of an OSC, UK operators are now able to take-off and land in areas they were previously unable to, and fly low enough to capture high resolution data. Obtaining this OSC however is not an easy feat and a high level of knowledge and operational understanding must be shown for a company to qualify.

With the development of regulation governing over flight of people and property, large parts of our cities are now able to be inspected and surveyed remotely using UAVs. The impetus has been placed on operators not to recklessly endanger people or property and there is even greater pressure on the pilot to be sure that a flight can be made safely.

By moving UAV operations from the countryside into the cities, a whole new range of risks have been introduced to operations. Using the 3P philosophy, it is possible for these risks to be reduced to acceptable levels. It is now the responsibility of the industry to do this in a professional manner so that the future of our industry is protected


[1] A. Welch, A cost-benefit analysis of Amazon Prime Air, Tennessee: University of Tennessee at Chattanooga, 2015.

[2] CAA, "Small Unmanned Aircraft: Congested Areas Operating Safety Case (CAOSC)," 12 November 2014. [Online]. Available: http://publicapps.caa.co.uk/docs/33/InformationNotice2014184.pdf . [Accessed 3 February 2017].

[3] CAA, "Unmanned Aircraft System Operations in UK Airspace – Guidance," 31 March 2015. [Online]. Available: http://publicapps.caa.co.uk/docs/33/CAP%20722%20Sixth%20Edition%20March%202015.pdf . [Accessed 3 February 2017].

[4] A. Moss, "Precisionhawk's FAA pathfinder phase 1 EVLOS report," 25 July 2016. [Online]. Available: http://www.precisionhawk.com/media/topic/precisionhawk-release-phase-1/ . [Accessed 3 February 2017].

[5] CAA, "Air Navigation Order," 10 October 2016. [Online]. Available: http://publicapps.caa.co.uk/docs/33/CAP393Ed5Am1_OCT2016_BOOKMARK.pdf . [Accessed 30 April 2017].