Building the AK-47 equivalent robotic drone

Former Boeing engineer Tilak Dissanayake has undertaken his most ambitious project yet – to build a cheap, mass-produced robotic drone to run long-distance air cargo operations

Tilak Dissanayake, a qualified engineer, is designing a low-cost, environmentally friendly robotic air cargo network – and he hopes to mass produce these robotic vehicles cheaply.

Unmanned aerial vehicles (UAVs) or drones have acquired a lethal reputation due to their wide deployment to fight who the US considers its enemies. Pilots on the ground – often thousands of miles away – control these UAVs through connections over a satellite net-work. These military drones fly way above the altitudes of commercial jets. When they unleash their guided missiles, people on the ground usually can’t figure where the bombs fell from.

But drones can be put to better use. The robotic air cargo concept is already somewhat followed by Google X’s Project Wing, which air-drops a payload of just three pounds, and Amazon’s Prime Air, which delivers five-pound packages to addresses within 10 miles of an Amazon center. The closest is DHL’s “parcelcopter” research project, launched in December 2013, which uses a quadcopter to deliver small parcels to the German island of Juist, a sandbar island 12km into the North Sea from the German coast. Deliveries include medication and other “urgent” goods.

CHT_7980Drone technology is pretty advanced. Dissanayake, because of his long aviation experience and general interest, knows this. He is not trying to reinvent the drone, but wants to make a technology leap similar to the one achieved by the Avtomat Kalashnikova (AK- 47), a weapon born at a different stage in the world’s violent past. The simple and reliable automatic rifle gained popularity as one that could be manufactured quickly and cheaply using mass production methods that were state-of-the-art in the Soviet Union during the late 1940s. Dissanayake wants to build a cheap, reliable, heavy lifting drone that can also fly itself to pre-set locations. Dissanayake – a bald, opinionated, dynamic retiree – has a bunch of daunting challenges to overcome in achieving this.

His primary focus is on the cost – the idea is to mass produce these drones so every courier company can afford and operate them. He foresees a future where tens of thousands of such drones will be criss-crossing the skies transporting everything from fresh produce to markets to necessary supplies.

Most cargo planes are standard airplanes converted to carry cargo rather than passengers. These planes need a commercial-sized airport to land and unload. The cost of operation is hefty and few private companies can afford to operate them. Sending cargo by these planes is also expensive.

“It is the AK-47 of airplanes. Very rugged, not sophisticated, but it works. It’s also inexpensive”

Dissanayake’s courier hub air robot (CHAR) is a robotically controlled cargo drone that will act as “a trishaw of the skies”. A standard-sized trishaw is 2.5m in length, 1.5m wide and weighs 325kg. The drone is 3m long, has a 2.5m wingspan and weighs just 300kg. It can reach a speed of 200km per hour, almost as fast as a military helicopter, and can carry a 20kg payload a distance of 50-60km on a full tank, almost the distance from Colombo to Marawila. The drone lands on its tail, much like a 1950s airplane and Google X’s Project Wing. The tail-sitter design and the drone’s size allows for vertical takeoff and landing from a 5m diameter pad. A cargo port for these drones can be just 10x10m, and located anywhere from the middle of a paddy field to a backyard.

“It is the AK-47 of airplanes. Very rugged, not sophisticated, but it works. It’s also inexpensive,” says Dissanayake who is designing the drone with his two young apprentices – Arjuna Mendis, a Star Wars-crazy electronic engineer who likes designing and building, and Manahara Manatunga, an aerospace engineer obsessed with World War II.

The AK-47 is a semi-automatic and automatic, gas-operated, 7.62×39mm assault rifle developed in the Soviet Union by tank mechanic Mikhail Kalashnikov during World War II. Having seen fellow soldiers and citizens gunned down, he wanted to ensure that it wouldn’t happen again. His design, the AK-47 came too late for the former Soviet Union’s (Russia) war against Nazi Germany, but went down in history as one of the most successful weapons. Today, for every 90 people in the world, there is one AK-47.

Dissanayake is positive about the project’s feasibility in Sri Lanka mainly because of the island’s uncrowded airspace

Dissanayake’s drone will be made of low-cost and strong aluminium strategically placed across moulded plastic wings. This is where the cargo will be stored. Even though carbon fibre would be a natural choice for an aircraft because it’s light and strong, Dissanayake feels the material and fabrication costs cannot be justified for a plane with such a low target retail price.

He uses components that already exist to keep the cost low, as well as sourcing material locally. Dissanayake hopes this will spawn a new industry in Sri Lanka with opportunities for small and medium businesses along the value chain.

To keep the cost low, they designed a fossil fuel engine for long-haul trips, which otherwise requires several battery swaps on the battery-operated version. Mendis designed the motor, of which a prototype was made in China.

Although launching with the battery-powered version, the fossil fuel engine will be used for the long range version to carry bigger loads of around 750kg. This is also because batteries have a very low specific energy capacity – with diesel at around 12,300 watt hours/kg, while the best lithium polymer battery is 230 watt hours/kg.

transparent_isometric_wo_linesManatunga is working on redesigning the flow geometry of a 1930s two-stroke, homogeneous charge, diesel engine. Dissanayake explains that, once completed, the high-power,
600cc, twin-cylinder engine will have less than 30 unique parts with only two moving parts, ensuring low manufacturing and maintenance costs. “We think we can retail price it locally at around $1,00-2,000 a unit, because we are trying to keep the whole drone at about $12,000-$15,000.”

The drone’s low-cost model also makes it feasible to be mass produced in developing countries. The reason Dissanayake is positive about the project in Sri Lanka is mainly because of our uncrowded airspace. “Developing countries have an enormous edge over cargo transport more than developed countries because the latter has problems with setting up segregated airways as the airspace is so crowded, with a huge general aviation populace that flies on VFR,” he says. The highest accident rate in the US is in general aviation and the safest is commercial traffic, which is under air control.

Dissanayake’s second challenge it to enable these drones to fly themselves. These drones are piloted robotically. This will enable mass production and eliminate security threats. “The human controllers will only have their hand on a big red button in case of emergencies,” he explains of the plane’s minimal interaction with ground control. According to Dissanayake, this is a starting step to make people comfortable that there is a human in the loop. Even the first motor car had a guy walking in front of it with a red flag to warn oncoming horse-drawn carriages of the day. As time goes on, these drones will become fully robotic.

The drone flies similarly to a quadcopter, a popular robotically controlled device flown as a hobby or recreationally. In Sri Lanka, quadcopters are popularly used for aerial photography. Dissanayake’s drone is also lifted and driven by four rotors or vertically oriented propellers (a type of fan that transmits power) placed two at the front and two at the back (see image). The drone is modelled after a World War II Mosquito light bomber created by the de Havilland Aircraft Company.

As they operate robotically, the drones will fly on segregated airways. A 300m wide and 200m tall airway will act as a virtual conveyor belt in the sky. The drones will fly steadily inside this virtual “box”, with a 2km separation between them. A 150m wide by 200m tall shoulder will allow them to temporarily jump out of the box in case of an emergency, much like a car stopping on the shoulder of Sri Lanka’s Southern Highway. The triple lane airways (in both directions) allow more planes in the sky at one time. It will be like setting up container shipping lanes in the sky. Dissanayake envisions over 5,000 planes in the air at one time as flight paths evolve, and 2-4 million worldwide.

The CAA’s role is that of a certification authority rather than a supporter to ensure that the robotic air cargo network is in compliance with all international civil aviation rules and regulations

The third challenge is to ensure safety. Commercial aircrafts fly at over 12,000 feet above ground level. The cargo drones will fly much lower. They have to fly 300m above the highest point on its flight path in flat territory and 600m above in hilly terrain. For example, the highest point from Colombo to Ambepussa is around 80m; so the drone will have to fly at 400m above ground level.

Helicopters are the problem that Dissanayake needs to sort out, as they fly on visual flight rules (VFR) as opposed to instrument flight rules (IFR). In VFR, the pilot is responsible for sensing and avoiding other traffic, so they have to see in front and around their chopper to navigate. IFR, which is used by commercial planes, is when the plane is under air traffic control, which manages air traffic and provides separation assurance. However, the robotic cargo drone’s segregated airways will show up on air navigation maps so helicopters can fly over or under them.

According to Dissanayake, the only reason for a drone to crash is if it is hit by a bird. The engineers are including everything from flashing lights to cameras needed to alert birds. But given the steady flying pattern, after a couple of hits, the birds will change their flying path, he predicts.

The only other thing that can bring this drone down is to shoot it with a stinger missile. “If you can buy a stinger missile, you might as well keep that money rather than shooting one of these down,” he jokes.

In an emergency, the drones are equipped with an automatic flight termination system and manual abort capability, which will deploy a parachute for safe landing.

1The biggest challenge is getting certification and funding for the project. “It’s a regulatory battle rather than a technical battle,” Dissanayake says, as most Sri Lankans are sceptical without a prototype.

He recently presented his project plan and business case to the chairman of the Civil Aviation Authority (CAA), Ananda Wimalasena, who was willing to take the project to the Ministry of Defence and the Air Force. The CAA’s role is that of a certification authority rather than a supporter to ensure that the robotic air cargo network is in compliance with all international civil aviation rules and regulations.

A standard-sized cupboard holds stacks and stacks of manuals on aviation and flight regulations like acceptable rules for compliance for engines that Mendis and Manatunga peruse daily as background work. Approval for the regulatory framework with the CAA is mandatory before launching the program.

“It is an absolutely daunting task, but a journey of a thousand miles begins with a single step. This is that step,” Dissanayake says.

On the plus side, certification creates a barrier to entry. A type certificate and approval, which is a patent for the product worldwide, will ensure that the first mover stays ahead of follow on, copycat products. The certifying authority provides the guardianship for your product and protects your investment. Dissanayake is sure the competition will follow, but the idea is to create a drone so affordable that it will be years before the competition can catch up. The team will also open the segregated airways to other aircrafts that comply with their standards, as they won’t be able to supply the world’s demand once the concept takes off.

After the concept initiation and certification with CAA, the next step is to find an investor and get funded. Thereafter, it will take three to four years for commercial operation.

The idea is to mass produce these drones so every courier company can afford and operate them

Dissanayake estimates the project as a $50-100 million investment. The main funding will be to pay the engineering team, purchase the necessary tools, and for flight tests and stimulation software. Although there is plenty of foreign investment available, he is an admitted nationalist keen on finding a local investor. “GNI (gross national income) is what helps us get out of the middle-income trap. So we have to be a Sri Lankan-owned company,” he explains.

Dissanayake is a self-proclaimed abundance theory follower, believing that you will find what you need because there is plenty of everything (resources) available. He lives by aerospace engineer Theodore von Kármán’s saying “Scientists discover the world that exists; engineers create the world that never was”.

An airplane nut from childhood days, he qualified as a mechanical engineer in 1971 from the University in Hawaii. He spent 12 years in the commercial and the defence sectors of The Boeing Company and seven years at Xerox, working in artificial intelligence applications. After short stints in Silicon Valley and a consulting startup company, he returned to Sri Lanka in 2001.

Although working full-time on his robotic air network project for the past three years, Dissanayake is constantly on the lookout to enhance the lives of Sri Lankans. “I love to think about problems, and actually do something about it,” he said in his recent talk at TEDx Colombo. Sri Lanka’s GDP of $75 billion lies mostly on the backs of our women, mothers and sisters slaving in the Middle East. He aspires to change this. He also sees a bright future for Sri Lanka’s 2,000+ engineering graduates a year, and wants to help them design products that the world wants to buy by the millions.