The designer who built a drone to save lives

We speak to Keenan Wyrobek, the founder of Zipline about designing an autonomous drone system which has revolutionised the way blood and medicine are transported in places which are hard to access.

Zipline, a drone delivery service bringing blood and medicine by air to areas which can’t be reached by road, is a social design endeavour that has not just changed, but saved lives.

At its heart, the creation of Zipline is a story of government and private enterprise working seamlessly, a strong concept and a solution to a very real problem based on an obsessive design and test programme.

Many of Rwanda’s towns and villages are not fully serviceable by roads, at least not in an emergency. With roads in disrepair or at the mercy of flooding it’s not difficult to imagine the problems involved in transporting blood and medical supplies.

In 2014 the Zipline system was set up to deliver these items by aeroplane drone as part of a targeted, sophisticated network. While the finer details are complex the idea is easy to visualise.

Two distribution centres with the capacity to serve 12 million people and 450 health facilities receive messages requesting blood or medical supplies. From here a drone is packed with a box to carry them and given coordinates. It’s then catapulted into the air before flying autonomously to its destination and jettisoning its cargo by parachute. This floats down to the ground where it’s collected by a medical professional at a designated hospital or health centre before the drone completes its return trip and is “caught” by a hook on a wire back at the control centre in a very precise act of aeronautical acrobatics.

Within 30 minutes almost any part of Rwanda can be reached. It’s not just the impassable roads which can be bypassed. Parts of the Virunga volcanic mountain range in the North West can be negotiated as well as Rwanda’s many rivers in the Congo and Nile basins.

Each distribution centre can service an 80km service radius and deliver a 1.8kg payload. Some 500 deliveries a day can be made. The battery-powered drones cover 40,000 km every week and if you want that put into perspective, it’s the equivalent of circumnavigating the planet.

We first met Zipline co-founder Keenan Wyrobek back in February when he was speaking in Cape Town, South Africa at the Design Indaba conference.

At the time he was already able to demonstrate the tangible difference the project was making and when we caught up with him this week we learnt how the Rwanda endeavour has propelled the project into new territories.

Testing and refinement of Zipline all takes place in California, US. “We don’t want to make mistakes in Rwanda” says Wyrobek, who has a Silicon Valley background.

He is perhaps best known for being the creator of the open source Robot Operating System (ROS), which is widely used to develop robots.

Many of his pre-zipline projects either “scaled wildly but didn’t matter, or I thought they mattered but I could never get them to go anywhere,” says Wyrobek.

Zipline presented a watershed moment in terms of new working methods. “I said I wouldn’t go and work on something until I had actual customers who wanted it and until I understood the problem”. This led to him meeting his co-founder William Hetzler who was looking to tackle a project in a similar way.

“We kept hearing about these big vaccination campaigns that Bill Gates runs or why World Health Organisation initiatives were failing because of logistics. Plus we both had public health people in our families who said you have to look at this thing [blood supply],” he adds.

A research stage was begun, looking at how blood and medicine could be transported in hard to reach places and then things started to move quickly.

Taking pencil sketches to presidents

“Within four months of traveling and looking into it by getting to know the problem we were meeting presidents of countries when we only had a deck of pencil sketches. That’s how we knew this was not a little thing. They wanted the problem solved,” Wyrobeck says.

What he soon found – particularly in the case of the Rwandan government – was that this keen interest meant a willingness to be highly collaborative from  central government down to a local level.

“Unlike many other governments they were amazing at using data to make good decisions. Right at the beginning we were told ‘These are our economists and our health systems, go crunch the numbers with them’”, he adds.

This asks a lot of broader questions about governments’ willingness to work with designers to solve problems. “Honestly, it’s embarrassing,” says Wyrobek. “In the US I’ve never seen a government meeting that went that way. It’s always more about politics, whose district you’re helping and so on, instead of just asking is this the right solution to the problem and do the numbers look good”.

This is one of the reasons that Rwanda was the starting partner – “They said ‘How quickly can we get this done,’” says Wyrobek.

At this point the real work started as the Zipline team began to create working models before improving reliability and safety. The design, engineering and testing that goes into this is “insanely complicated” Wyrobeck offers, particularly because the drones are bespoke.

“I can tell you what every part weighs on our drones to the gram. That’s what you need to know when you design a plane like this,” he says.

While Hezler was working on business development Wyrobeck had convinced himself that the product development would be relatively straightforward and began by looking for an off the shelf drone – possibly something which had been designed for another application – to modify it.

“It was a sobering moment when after six months talking to everybody who makes drones I realised we had to do it ourselves. The ones I’d encountered either couldn’t do the delivery, couldn’t handle the range or couldn’t manage the weather,” he says.

Accordingly everything was designed from scratch and a team of some 40 designers was put together to take on this challenge, comprising aeronautics engineers, robotics engineers and systems engineers – all poached from companies including SpaceX, Google and Toyota.

“Testing is what we do for a living”

There was a lot to work out. How to launch the planes. How to fly through a storm when close to the ground. How the parachute delivery system would work. Could the plane deploy its own parachute if it gets into difficulty?

The process was forensic. Taking just this last conundrum – the plane deploying its own parachute – the design team could see it needed to be effective if the plane was compromised from any angle, like if it was hit by a bird. For this purpose, confusingly, an actual zipline was created to test Zipline, so that it could deploy its parachute in various scenarios at 100km/h. “Testing is basically what we do for a living,” Wyrobeck says.

Another demonstration of this obsessiveness is the weather testing. If a designer wants to change a component on a plane, like a tail fin, it’s tested to destruction in a shipping container which can simulate humidity, rain, wind and turbulence on a sliding scale of ferocity. “If it survives for a couple of days, we put it out,” he adds.

It took two years from the beginning of the project to making the first flights in Rwanda. This first generation system is being constantly tweaked and will naturally inform the development of systems in other countries.

While the design was being honed the whole enterprise was being invested in by superangels and venture capitalists. With all this design pedigree and a large, clearly identifiable healthcare logistics market in rural areas and smaller cities, there was a strong business case.

Indeed Zipline has been devised as a scalable business. Not a charity, social enterprise or pro bono project. Although it is pro Bono. The U2 front man joined the Zipline board in September and is now an advocate.

The business relationship with the Rwandan Government is a very practical one. At the top of the contract it says: ‘Courier Service Provider to the Government of Rwanda’, according to Wyrobeck who says, “We’re the logistics arm of the public health system; if we can solve the problem – access to supplies – then we will be a successful start-up business.”

It helps that Zipline is a responsible employer. Everyone trained to work for the company in Rwanda is Rwandan. Today that means around 40 people and the benefits of this kind of model are easy to see at a central Government and local level.

Despite the convivial relationship with the Rwandan Government, working in the skies is fraught with regulatory and practical challenges. And with good reason. The International Civil Aviation Organisation’s Eastern and Southern Africa office was set up in 1983 to improve safety in that region of Africa. The ICAO standardises and regulates air travel and Zipline had to fall in line with these rules.

Zipline works with all sorts of regulatory bodies formally and informally including the Rwanda Airport Authority and the Federal Aviation Administration, which it is working alongside to create a set of standards for drone delivery systems in general. It sounds like working with regulators can be slow going though – “Don’t bet on a regulator’s timekeeping. That’s one thing I’ve learnt”, Wyrobeck says.

The company is now poised for expansion. The business has been a demonstrative success in Rwanda and one of the biggest success stories is the effect on maternal mortality rates.

Indeed there are three hospitals in particular which Zipline serves in Rwanda – Shyira District Hospital in Nyabihu District, Kacyiru District Hospital in Kigali and Ruli District Hospital in Gakenke District – where there were no maternal childbirth deaths during a seven month test period.

While this can’t be attributed to Zipline directly, there are qualitative examples showing that its influence was a major contributing factor.

Shyira District Hospital director general Maj. Dr. Emmanuel Kayitare, told Rwandan newspaper The New Times that “the use of drones for blood deliveries where it is required as an emergency” was making a difference in the case of transfusions for caesareans, hemoglobin conditions and spontaneous delivery.

Zipline was an effective part of a front line service at these hospitals which also included a new referral scheme from health centres and better communication with patients.

Although Design Week was unable to access maternal death rate information on these hospitals prior to Zipline’s contribution, we do know that maternal death rates are a very real problem in Rwanda.

In 2015 a report by the World Health Organisation showed that a woman in Rwanda had a one in 85 chance of dying from a pregnancy related complication in her lifetime, compared with a woman in Norway who had a 1 in 11,500 chance of dying from the same complications.

The wider context points to childbirth as an increasingly important focus for Zipline as the population of youth in sub-Saharan Africa will be the fastest growing in the world and could account for more than 400 million people by 2050, which among other things, means a lot of people are about to be born.

Where next?

Although Zipline hasn’t limited itself to Africa it is ruling out large Metropolitan cities, focusing instead on remote areas and small cities. In April Zipline started up its first of four distribution centres in Ghana. The national project will soon be able to serve 2,000 health facilities and a population of 12 million people across the country.

In 2020 Zipline will be launching its service in India, which could serve up to 120 million people over the next several years. The next generation of Zipline is expected to be ten times more efficient and serve around 100 times more people.

The scale of ambition is both staggering and quantifiable. Tens of thousands of lives are expected to be saved and 700 million people served, all underpinned by commercial partnerships with the governments of Maharashtra, Ghana and Rwanda.

There are even plans to set up in the most remote parts of the US, a task which will presumably encounter more political and bureaucratic hurdles than the other locations.

We often find that service design can be intangible but Zipline is a great example of this type of design as it’s a complete service with real world physical touchpoints. However the planes themselves make up about 10% of the complexity of the service according to Wyrobeck.

While scenes of soaring planes carrying boxes of blood autonomously captures the imagination, this feat is the output. The main design challenge is much more sobering.

Wyrobeck says “The biggest challenges of setting up a national drone delivery service involved integrating with a public health supply chain and building all the infrastructure and equipment necessary to do 150 flights per day, reliably and routinely.”

It puts the commercial drone delivery research and development work that the likes of Amazon are currently experimenting with into perspective as well. We often see prototypes of these projects or rendered CGI visions of the future showing skies full of hovering drones but this kind of system is already a reality and it’s blood rather than gadgets and homewares which is being delivered.

Latest articles