Canada’s Moon Shot aka Forget About 5th Gen Fighters

This is a Friday thought piece followup to my commentary on Canada’s anti-innovation cultural norm, and is mostly outside of my realm of expertise, so expect little and read with a sense of humor.

While I’m out of my depth on the military side, I’m fairly knowledgeable of most of the technology considerations — embedded, hardened real-time systems, communications systems, machine learning and “AI”, advances in image recognition — and have a good history of accurately predicting technology trend outcomes aside from HD-DVD but who remembers that anyways, right?.


Canada should forget about buying a 5th generation fighter and initiate a large scale effort with the aerospace and technology industry to build a low-cost drone to fulfill the sort of missions Canada will foreseeably take part in, leveraging industry solutions and engaging allies, but not being beholden or a second tier behind anyone.

The technology and capability is ready, and it’s the right solution for the needs.

Canada’s Aging F-18 Fleet

Canada is currently in a procurement process to replace a fleet of aged (C)F-18 fighters first put into service in the early 1980s. A fleet that originally started with 138 aircraft, though substantially fewer remain airworthy (estimated at about 80).

I spent my adolescence pouring over military specifications and books, dreaming of being a fighter pilot (the Top Gun era and all), but that dream was abandoned given the anemic state of Canada’s air power.

While Canada has been a partner in the F-35 program since the beginning, there have been wavering commitments given the rising costs and time slips, the number of aircraft that might be purchased dropping below the already anemic 65 unit plan.

Sixty-five aircraft for a country of 10 million square kilometers, much of it remote. Concern has also been raised about it being a single-engine aircraft operating in Canada’s arctic regions where there are no airports or nearby rescue crews.

The pilot will likely survive an engine failure, barring a polar bear encounter, but it is a sure loss of an airframe (this is an aircraft that is pegged to have a $200M CAD purchase price tag, and an almost $700M lifetime cost per unit).

If you have 1700 F35s, a loss of aircraft isn’t a huge deal. It is a big problem when you have 65 or fewer.

The US refuses to share the source code that runs the aircraft, which ultimately should make it a non-starter from the outset: While the two countries are interconnected and almost indistinguishable, there are theoretical scenarios where Canada may cooperate with the UK, for instance, on a venture that the US may not agree with, etc. The notion that someone else ultimately retains ultimate control over your military hardware should be an elimination factor. No country should ever cede its sovereignty in such a way.

5th Generation Fighters Will Be The Last

There will be no manned follow-up to the F22 and F35. The future is unmanned, and I would argue that 5th generation fighters will have an extraordinary short window where they’ll be viable solutions, and are in some ways obsolete on delivery, mostly poised to be airshow performers.

They’re way too expensive, to start.

These fighters are prohibitively expensive because a fundamental, driving goal in their design is that the pilot cannot be lost.

Putting aside the tragedy of loss of life, pilots are expensive to train, and are in some ways the military elite, if not symbolic. Pilots often operate behind enemy lines (most engagements utilize them as offensive weapons), and add considerable political cost to military pursuits if a pilot is captured, or a body can’t be recovered, in enemy territory.

Having hardware that allow you to operate with reduced concern about the enemy forcing mock confessions from your pilots on YouTube is worth billions. Which is why an F22 is a $200M USD aircraft, yet a lot of ground troops can’t get body armor.

Consider that the F16 (a gorgeous design) was originally built with no radar, very limited avionics, and was considered an inexpensive day fighter. The goal was power in numbers: Build lots of relatively minimalist aircraft cheap and just swarm the air, originally planning against a massive Soviet invasion. It wasn’t long before it started to get feature creep (“Can the congressman explain why he’s willing to let our boys be shot down in this subpar…”), becoming an all-weather, all-purpose, air-superiority fighter/bomber.

The Limitations of the Meat Bag

The meat bag in the cockpit doesn’t just completely change survivability demands (overriding pragmatic considerations like “if we can build them for 1/3rd the price, and they’re 1/2 as survivable, let’s build 3x as many of the cheap ones“), they also represent added complexities: Environmental systems. The physical limits of the human body. Design for visibility. The space required to comfortably and safely hold a human and all of the safety gear around them (e.g. ejection seat). The need for the aircraft to make it back to friendly territory regardless. Every risk factor needs to be gauged based upon keeping that human alive.

There may be a situation where from a practical perspective it’s worth sacrificing an aircraft to achieve a goal (a high priority target is available but will use the remainder of fuel), but this is a non-starter with a human in the cockpit, at least outside of WW2 kamikaze pilots.

Now build all response plans around those humans. Do you have people sitting on deck ready to take off at a moment’s notice? How quickly can you respond if a Tu-95 appears at the edges of radar limits, probing response times and simulating a nuclear run?

The advantages of unmanned aircraft are overwhelming.

Very high-G launch/recovery techniques? Why not. Who needs an airfield (a tactical weakness that you need to defend, and that has considerable orchestration overhead) when you can launch straight up, close to instantly. When you can recover via parachute, if you even recover at all. Remove the person, and open up the risk profile (failed launch? Big deal), and the parameters of what is possible improves dramatically.

This isn’t some future world but is a change that has been underway for decades. With each engagement the number of sorties carried out by aircraft has declined, being replaced by autonomous or remote control vehicles. Cruise missiles were the first volley in the first Gulf War, doing the high risk missions that attack aircraft were traditionally tasked with.

288 cruise missiles were fired in the first Gulf War.

725 cruise missiles were fired in the second Gulf War.

In a recent assault on ISIL the US fired 47 cruise missiles and dropped 36 smart bombs. Even when there was overwhelming and absolute air superiority against a foe with very limited capacity to respond, autonomous vehicles were selected.

Cruise missiles are expensive, though. Something like a million dollars each, versus $40,000 for a smart bomb. Which is why the US has increasingly depended upon UAVs as well — a relatively inexpensive airframe carrying an inexpensive payload.

The Role of Modern Aircraft

Which gets to the role that fighter aircraft really fill. In a modern era conflict they’re a taxi cab for shorter range weapons.

They bring air to air missiles within range of their target, hoisting a radar to a more useful height and perspective. They carry bombs to a location (altitude and distance) where they have enough potential energy to hit the target.

5th generation fighters do not do traditional recon, and usually operate entirely outside of visual range of their targets (whether on ground or in the air). They don’t do close air support.

They’re an electronic information gathering weapon sled. The weapons they deliver are often entirely autonomous after separating — that AMRAAM or Maverick or smart bomb are the master of their own destiny once they leave the launcher.

In non-combat they’re a show of force. When a Russian Bear comes edging towards airspace, they make their presence known, fly in tandem and do the charade, and everyone goes home and has a drink of vodka. If it was an actual high tension situation where the threat of a nuclear first strike was imminent, they would have unleashed weapons one hundred miles out.

What about any of this can’t be accomplished with automated devices? Right now, with available knowledge and software and technology?

The F35 program was started in the 1990s. The F22 program was started in the 1980s. They’re the end result of an obsolete set of inputs for a world that no longer exists.

Consider that the newest F22 computer performs about 10 billion operations per second. The Tegra PX2, a low-cost SoC destined for the self-driving car market, performs 24 trillion operations per second.

In every facet of the discussion we’re in a different world. Processing power, imaging and image recognition, sensors of every sort (I mentioned before, but not long ago gyroscopes were expensive, error prone physical devices).

So the Canadian military needs to occasionally go meet a Russian bomber at the edge of airspace just to play the cold war reminiscing game. Slap a jet engine on a basic airframe (yes, facetiously simplified) and send the drone off. Make them cheap enough and keep an endless procession of them flying routes around the Northern borders, carrying information gathering devices that give insight into the arctic that is currently missing.

But what about jamming!

This is the most prevalent argument against unmanned aircraft. Yet jamming would be a critical problem for our current manned aircraft (integrated battle nets fed by AWACS and ground radars, along with friend/foe tagging information. Even final approval to fire weapons — manned aircraft rely upon a constant connection to function), so military communications systems have been built to be largely jam resistant. This is doubly true when considering defensive purposes where your control of the land gives you a strong advantage in any radio signal battle.

But consider that missions are almost always pre-determined before launch, whether manned or unmanned (aircraft seldom loiter on station). If a nuclear first strike were deemed imminent, fighters wouldn’t intercept and tag along with a Bear, and instead would shoot it down far over the horizon. If it’s a normal peacetime intercept not only won’t there be jamming attempts (an act of war), a drone can intercept and tag along and feed back an enormous trove of up to date information, and more drones could be ready with more hostile directives if the first came to bad ends.

Extraordinarily little in a modern mission is planned or changed after the aircraft has left the base, and under that situation the drone would be in the same situation if it were fully jammed. Fly from here to there, and if you find anything airborne that isn’t squawking the right codes, shoot it down. Correlate with civilian flights (if not in a state of war where civilian flights stop happening), use an analysis engine to accurately identify the target, etc.

But of course that’s a worse case. In realistic situations it would be in constant encrypted and controlled communications.

Skynet, hacking, etc…

A significant concern, but the future is happening whether we’re a second-tier customer or an instigator. Might as well get some inside info on our eventual AI foes by helping to build them.

This Should Be Canada’s Moon Shot

The country is looking to spend a large amount of money on an already obsolete solution. It should use this opportunity to instead boost the Canadian technology and aerospace industry, especially in the face of potential upcoming US protectionism. Simple drones to fulfill basic tasks like monitoring and defending the arctic are achievable undertakings even over a tight time window, and tactical capabilities could be vastly improved with cheaper options.

Throw out everything you think about air power and start anew, because nothing is the same anymore.

Waiting for the industry to deliver the solution isn’t satisfactory. Canada’s needs are somewhat unique, and the people who would deliver this sort of solution generally have their hands so deep in the money pot they’re going to do nothing at all to upset what they already have going on until they have to, though a relatively small drone industry (largely the Israeli tech scene) is starting to force some hands.

Canada has the talent and the ingenuity, and if other allies wanted to come onboard so be it. The country should control its own destiny, instead of forever being an eventual customer of other people’s solutions.