Are We Ready for the Drone Future?

On a weekday in central London, a drone carrying blood samples lifts off from Guy’s Hospital and lands at St Thomas’ a few minutes later. By road, an urgent courier can take around half an hour. At Notting Hill Carnival in August 2025, a police drone with a thermal camera flew above a weekend crowd of around two million people, able to identify a single person from about 80 metres away.

The same technology, doing two very different jobs. A drone is simply an aircraft with no pilot on board. It is flown by someone on the ground with a remote controller, or by an onboard computer following a pre-programmed route. What it actually does is carry something, usually a camera, a sensor or a small package, to places that are slow, expensive or dangerous for people to reach. That is why drones now turn up in hospitals, on railways, over farms and building sites, at offshore wind farms and on the battlefield. Both of those flights in London are part of Britain’s drone future, and both arrived faster than most people noticed.

In July, we argued that the UK’s drone industry was about to explode, and that the thing holding it back wasn’t the drone. It was the law. Until drones can routinely fly beyond their pilot’s line of sight (BVLOS), most commercial uses don’t scale, and the government has set 2027 as the target to make that routine.

Three months on, a lot more is flying. Hospitals, police and offshore wind operators are running BVLOS flights, and on 26 October the head of the RAF said its future will rely on “autonomy, not flesh”. So the question has changed. It is no longer whether the drone future is coming, but whether we are ready for it: the rules, the people who will be watched by drones, and the people who will build and run them.

Here is what has changed, who is flying, who is setting the pace, who is being watched, and what to watch next.

Drones in brief

Most drones fall into three groups:

• Consumer quadcopters: the most common type, with four propellers and the ability to hover in place, used by hobbyists and photographers
• Commercial drones: larger, specialised models built to carry packages and medical samples, or to inspect infrastructure
• Military uncrewed aircraft: from cheap first-person-view drones to large, long-range systems for surveillance and security, flown without putting a pilot in danger

Drones sit mainly within the aerospace and defence industry, but they cut across many sectors:

• Defence and security: surveillance, reconnaissance and autonomous systems
• Energy and infrastructure: inspecting power lines, railways, wind turbines and pipelines
• Logistics and delivery: medical samples, parcels and warehouse stock checks
• Agriculture and forestry: precision farming, crop spraying and land mapping
• Construction and property: site surveys, 3D models and aerial photography

This article follows the sectors moving fastest in the UK right now: health, policing, energy, deliveries and defence. It also looks at the hundreds of thousands of drones in ordinary people’s hands.

What has changed since July

The drone future is already flying

Hospitals: from a trial to a network

In July, the NHS was our glimpse of the future. In London it is becoming the present. Apian, the healthcare drone company, has flown more than 13,000 clinical items across the capital since October 2024. Its first route, between Guy’s and St Thomas’, moved more than 6,000 samples in its first phase. Apian reported average cost savings of 28% and a 98.4% cut in carbon emissions per delivery compared with urgent courier vans. The CAA has approved continued flights on that route until February 2027.

The network is spreading. Since February 2026, Wing has flown samples from Nelson Health Centre in Raynes Park to St George’s Hospital in Tooting. In October, Matternet began carrying children’s pathology samples from Great Ormond Street Hospital, the third London trust to take part, and the Guy’s and St Thomas’ corridor was extended to the Synnovis hub laboratory at Blackfriars.

This is the model we described: not replacing ambulances, but finding the specific journeys where a small, urgent package moves faster by air than by road.

Railways: policing goes routine

The clearest sign of progress came in September, when British Transport Police became the first UK force to make BVLOS drone flights part of daily policing. Its drones live in drone-in-a-box stations across the rail network and are flown remotely from a control centre at London Bridge, connected over the EE mobile network using BT’s Drone SIM. Officers get live video of trespass, welfare concerns or disruption before anyone reaches the scene. Seven stations are running, with up to 14 planned over the next two years.

This is drones as infrastructure: no pilot standing in a field, a fleet run from one room, and a mobile network doing the work that used to need a line of sight.

Offshore wind: skipping the boat

In October, Glasgow start-up Airspection announced that a drone launched from the shore had flown entirely beyond visual line of sight to a working offshore wind farm in Liverpool Bay and back, a round trip of about 14 km. Inspection drones usually have to be taken out by boat. Not needing one is the point.

The flights were part of the CAA’s BVLOS sandbox, in which the regulator works with six projects to gather the safety evidence it needs. The others include the National Police Air Service, NATS over the North Sea, Project SATE in Orkney, Project Lifeline for medical deliveries and Amazon Prime Air.

Darlington: parcels in the garden

Amazon’s first drone deliveries outside the US are running from its Darlington fulfilment centre. Customers with a garden or driveway within 7.5 miles can receive items of up to 2.2 kg (5 lb). Reports describe the service as a CAA trial running to the end of 2026, limited to 10 flights an hour and 100 deliveries a day on weekdays.

Darlington also shows a part of the story we underplayed in July. The local council approved only a temporary structure with a single launchpad, and some residents have raised concerns about noise and privacy. Regulation decides whether drones can fly. Communities decide whether they are welcome.

Drones in everyone’s hands

Our July article was about industry. It missed the biggest fleet of all: the drones in ordinary people’s hands.

Ten years ago, drone crime in London was rare. Answering a London Assembly question in 2016, the Mayor reported that the Met had recorded 42 drone-related offences across the whole city in the two years to April 2016, too few to break down by borough, and had made five arrests. Today the CAA’s register lists around 607,000 active flyer IDs and almost 331,000 operator IDs across the UK, according to its figures to 31 May 2026. Since 1 January 2026, drones in the new UK1 to UK3 classes must also broadcast a Remote ID when they fly.

Over the landmarks

Most of that flying is harmless: sunrise shots over the Thames, holiday footage, a roof survey. Some of it is not. In 2015, a hobbyist became the first person to be prosecuted by the police and the Crown Prosecution Service over drone flights in London, after flying over the Houses of Parliament, Buckingham Palace, the Shard and football stadiums. He was fined and banned from flying drones for two years. In 2021, another pilot was fined more than £5,000 for flying over MI6’s headquarters.

Behind the music video

Drones have also changed how our culture is filmed. A shot that sweeps up above an estate has become a visual signature of UK drill and grime videos, as researchers of the scene have noted. Aerial footage that once needed a helicopter and a film crew’s budget now costs a few hundred pounds. That is a creative opening, and a route into paid filming work. But the rules on flying near people and over built-up areas apply whether the footage is for a music video or a holiday. And drill videos already draw close police attention: in 2018, Met complaints led YouTube to remove around 30 of them.

Over the prison wall

The sharpest misuse is over prison walls, where drones are used to fly in drugs, weapons and phones. Recorded drone incidents at prisons in England and Wales rose from 357 in 2021/22 to a record 1,712 in the year to March 2025, before falling to 1,497 in the year to March 2026. The prison service says part of the rise may reflect better reporting, while the Commons Justice Committee said in October 2025 that the official figures probably understate the problem.

The government has put £35 million into security at the most at-risk prisons, including heavy-duty window grilles, and £6.5 million into research on anti-drone technology.

This is the other side of the readiness question. The same cheap, capable drone that lets a young creative film their neighbourhood from the sky lets a gang drop a parcel at a cell window. The rules, the detection systems and the drone ID scheme described below all have to work for both.

Defence is setting the pace

Civilian drones are growing route by route. Defence is moving in thousands. In September, the British Army said it had bought more than 10,000 drones in a year through Project Rapstone, its fast-track buying programme, most of them for training. Soldiers learn to fly first-person-view drones by hand, without GPS, because the hardware changes so quickly that skilled pilots outlast any single product.

The money is now clearer. The Defence Investment Plan, published on 30 June 2026 with an extra £15 billion for defence, commits more than £5 billion over four years to what the government calls a “drone transformation” of the Armed Forces. For scale, the whole plan is worth £298 billion over four years, so drones are under 2% of it. That includes £650 million for low-cost, expendable autonomous systems, such as drones and uncrewed ground vehicles, for the Army, Commando Force and Special Forces. Alongside it, the plan sets out:

• £790 million for air, drone and missile defence, including more counter-drone systems
• Nearly £2 billion for a Digital Targeting Web, AI-backed software that links the Armed Forces together for faster decisions (no timeframe given)

The Army has also selected four industry teams (Anduril UK, BAE Systems, Tekever and Thales UK) to compete through Project NYX to develop uncrewed systems that work alongside its Apache helicopters.

“Autonomy, not flesh”

The RAF went further. In a speech at the Royal United Services Institute on 26 October, Air Chief Marshal Harv Smyth, the Chief of the Air Staff, set out a future in which every Typhoon fighter flies with four armed, AI-enabled drones under the £300 million Storm Fighter programme. The Guardian reported that the first tests are due next year, with the aim of having the drones in service by the end of the decade, in line with NATO’s call for allies to be ready to fight Russia by 2030.

Smyth was open that people are part of the reason. The RAF’s fully trained strength has fallen from 29,800 in 2020 to 27,850 today. In his picture of the future, humans move from flying every mission to supervising the systems and setting the guardrails they operate within.

The ethical line

Those guardrails are where the hardest question sits. AI has already been used to speed up target selection in Gaza and Iran, and they only mean something if a person remains accountable for every strike.

The UN has been clear about where it stands. In August 2026, UN Secretary-General António Guterres and Mirjana Spoljaric, President of the International Committee of the Red Cross, renewed their joint call for states to start negotiating a legally binding treaty on autonomous weapons. It would ban some systems outright and strictly limit the rest, with particular concern about machines that select and attack people on their own. The next test comes in November, when states meet in Geneva for the Seventh Review Conference of the Convention on Certain Conventional Weapons, after more than a decade of UN expert talks.

The UK is more cautious. It voted for the UN General Assembly’s December 2025 resolution on autonomous weapons, and says it opposes systems that operate without “context appropriate human involvement”. But it does not support negotiating a new treaty, preferring to rely on principles under existing international humanitarian law.

From the battlefield to everyday life

Defence setting the pace is not new. The war in Ukraine has shown how cheap, mass-produced drones can change a battlefield in months, and governments have moved money, procurement and regulation quickly to keep up. Many of the technologies we use every day began the same way. The internet grew out of ARPANET, a network funded by the US Department of Defense. GPS was built by the US military before it was opened up for civilian use. The jet engine was patented by an RAF officer, Frank Whittle, and the microwave oven came out of wartime radar research. The autonomy, computer vision, navigation and edge AI now being built for defence will find their way into hospitals, railways and farms.

As an innovation institute, we see both sides. The boom, and the coverage that comes with it, has sparked real interest in drone innovation, bringing investment, jobs and new companies. But technology shaped by the battlefield brings the battlefield’s priorities with it. Defence values speed and secrecy. Public services need transparency, consent and accountability. Whether that turns out for better or worse depends on who makes the choices.

The rules are still catching up

If defence is moving in thousands, the civilian rulebook is still moving page by page. On 30 July 2026, three weeks after our original article, the CAA published its Airspace Architecture Concept of Operations. It sets out how drones could share airspace with helicopters, light aircraft and airliners, moving away from today’s case-by-case approvals and fenced-off airspace. It names five building blocks, and they line up closely with the questions we asked in July:

• Electronic conspicuity, so drones and crewed aircraft can detect each other
• Detect-and-avoid, so a drone can get out of the way on its own
• Resilient command-and-control links between the drone and its remote pilot
• Digital traffic management to coordinate large numbers of flights
• Information sharing between operators, air traffic services and other airspace users

It is important to be clear about what this is. The document creates no new rules, requirements or permissions. It is a direction of travel. The detailed rules for drone traffic management went out to three CAA consultations that closed on 28 August 2026, and the CAA has said it will publish a response.

Money is following, slowly. In May 2026 the government announced £46.5 million for drones and flying taxis, including about £20.5 million for what the Security Minister called “a numberplate system for the skies”. It is a drone ID system that broadcasts a drone’s identity and location as it flies, so police can pick out illegal or nuisance flights. Making legitimate operators easy to identify is part of what earns everyone else permission to fly.

Meanwhile, almost every civilian flight in this article runs on a temporary approval. The Guy’s and St Thomas’ route runs to February 2027, Amazon’s trial reportedly to the end of 2026, and Great Ormond Street is on a six-month approval. Each was negotiated route by route. That is progress, but it is not yet routine.

Who is watching, and who is watched

For many of us, the first drone we notice won’t be carrying blood samples or inspecting a wind turbine. It will be a camera in the sky above a celebration of our own culture.

At Notting Hill Carnival in August 2025, British Transport Police, the same force now flying BVLOS on the railways, deployed a dedicated drone team for the first time. The BBC reported that the drones carry four cameras, including thermal imaging and a powerful zoom. A laser rangefinder can tag a location more than a kilometre away and send its GPS coordinates to officers on the ground.

Carnival is Europe’s largest street party, and the Met has said a crowd crush is the biggest risk it plans for. Seeing a crowd from above can save lives. But the same year, the Met also brought live facial recognition to Carnival, and civil rights groups argue that the technology is less accurate for women and people of colour. At Carnival’s 60th anniversary this August, the Met expanded its use of facial recognition, reporting 214 alerts and 121 arrests.

This is why drones aren’t only an aviation story. Who gets watched, how footage is used and how long it is kept are design decisions. They are made in safety cases, data policies and consultations, usually long before anyone notices a drone overhead.

We have seen this before: CCTV

In the mid-1990s, the Home Office paid for town-centre CCTV to prevent crime. Its CCTV Challenge Competition funded 585 schemes between 1994 and 1999, and for a period CCTV took most of the government’s crime prevention budget. The cameras were sold as a way to make high streets safer. Over time they became a national network, later joined by number-plate recognition and now live facial recognition.

Birmingham shows how that drift can land on particular communities. In 2010, Project Champion installed more than 200 cameras, many of them number-plate readers, around the mainly Muslim areas of Sparkhill and Washwood Heath. Councillors had been told the scheme would tackle crime and anti-social behaviour, but it was paid for from a counter-terrorism fund. A council scrutiny inquiry found no justification and criticised the lack of consultation. The Chief Constable apologised, and the cameras came down.

Technologists call this function creep: a tool built for one purpose quietly takes on another. A drone is a camera that can move, follow and see in the dark. The lesson from CCTV is that limits on purpose are easiest to set at the start, before the network exists.

Would drones pass the Townhouse test?

So how ready is the drone future, really? At the UKBT Institute we use the Practitioner Readiness Level (PRL) to test innovation. Before anyone asks “can we build it?” or “will someone pay?”, it asks “should we build it?” Twelve voices validate every idea in the Townhouse, each with one question only they can answer. Eleven of the twelve must say yes. Any red flag means pivot, kill, or an explicit, recorded decision to accept the risk.

So we put drones through it. This is our own assessment, based on the evidence above:

The pattern is clear. The voices that buy and build drones are saying yes. The voices that have to live with them have hardly been heard. In Townhouse terms, a red flag needs a recorded decision to accept the risk. Too much of the drone rollout is accepting that risk without anyone writing it down.

Where the jobs are

If the voices are missing, the fix starts with skills. Industry recruiters describe a two-speed market: plenty of people who have passed a basic drone exam, and a shortage of people with specialist skills. UAV Jobs lists BVLOS flight planners and safety-case engineers at £55,000 to £75,000, compared with £28,000 to £45,000 for a general commercial remote pilot. The CAA is also moving to a new tiered Remote Pilot Competency framework. On the defence side, the government expects its investment plan to create nearly 60,000 extra UK industry jobs by the end of the decade, though that figure covers all of defence, not just drones.

Look again at the picks-and-shovels table from our July article: fleet management, detect-and-avoid, airspace management, edge AI, cybersecurity, navigation resilience. Every development in this article depends on them. The long-term value sits in the skills, software and systems around the aircraft, and many of those jobs do not involve flying a drone at all.

So, are we ready?

Partly. It depends who “we” are.

• The technology is ready. Hospitals, police and energy companies are flying BVLOS every week, and defence is buying drones by the thousand.
• The rules are not, yet. The regulator has drawn the map, but most flights still run on temporary, route-by-route approvals, and enforcement is still catching up with misuse, from landmarks to prison walls.
• Communities have barely been asked. From Darlington’s launchpad to the cameras over Carnival, people are meeting drones before they have had much say in how they are used.
• The workforce is the gap. The specialist roles are scarce, and the RAF is betting on autonomy partly because it cannot recruit enough people.

The next 10 to 20 years

The official milestones are already set: routine BVLOS flights by 2027, flying taxi regulation from 2028, AI-enabled RAF drones in service by 2030, and a government estimate that drones and advanced air mobility could add up to £103 billion to the UK economy by 2050.

From here, we can see two futures. In one, drones become infrastructure as ordinary as streetlights, moving blood samples, inspecting railways and reaching emergencies first, with clear rules on what they record and why. In the other, a network of flying cameras slowly becomes the default way public spaces are watched, from the high street to Carnival, without anyone having decided that it should.

Go back to the two drones at the start: one carrying blood across London, one watching a crowd at Carnival. They are the same technology. Which one defines Britain’s drone future depends on who is in the room while the rules are still being written, and on whether the people who design, run and govern these systems over the next few years include us. That is a skills question, and it is one our community can answer. Never skip a voice.

Four dates to watch

• The budget, Wednesday 28 October. Prime Minister Andy Burnham has said Chancellor John Healey’s first budget will set out a path to spending 3% of national income on defence by 2030. It also has to find £4.7 billion that was left unfunded when the defence plan was announced in June. The Office for Budget Responsibility estimates that reaching 3% would cost an extra £17.3 billion in 2029–30. Healey had earlier said the timetable would come at next year’s spending review, so watch whether a firm date appears, how much of the new money reaches drones and autonomy, and whether anything goes to the civilian side, such as the regulator and NHS drone routes.
• Geneva, 16 to 20 November. States decide at the Convention on Certain Conventional Weapons Review Conference whether to start negotiating a treaty on autonomous weapons, and where the UK lands.
• The CAA’s response to the drone traffic management consultations, which will show whether the July blueprint is turning into rules.
• Early 2027, when the temporary approvals run out, especially the Guy’s and St Thomas’ route in February. Renewed, made permanent or stopped, the answer will show how close routine BVLOS really is.

We’ll check back in the new year.