Fighting Ebola With Robots And An App Called Jedi

In West Africa, a crisis group and two tech companies are building a high-tech clinic to bring better communication and data to the fight.

Fighting Ebola With Robots And An App Called Jedi
[Monrovia photos: courtesy of the International Rescue Committee]

Monrovia—Sometime soon a slender robot that looks like Casper the ghost and works like Skype on wheels may visit the bedside of an Ebola patient in West Africa, as a doctor nearby instantly transmits data to other researchers over a portable Wi-Fi network.


For now, the bot is traveling in a suitcase with Debbie Theobald, the CEO of Cambridge, Mass.-based Vecna Technology, who came to Liberia (where Ebola is still raging) in late November to help bring an Ebola treatment unit into the 21st—or 22nd—century.


Even Theobald doesn’t seem quite sure of the bot’s destiny in the clinics: “I must admit I am curious to see what other benefits the caregivers can see from a robot that can go and talk to patients without having to suit up in PPEs,” or personal protection equipment, which can take about 20 minutes to put on.

The robot, along with new medical software hacked together from a system already in use in veterans hospitals in the U.S., is a vanguard of the technologies that are quietly being deployed in the next round of the Ebola fight.

The clinic is the first to be designed and built by the International Rescue Committee, and it doesn’t look like many others here. Among the familiar tents, beds, orange fences, and people in plastic suits, providers will carry tablet computers that run on the new EMR system.

Upgrading Veteran Software

It took one month for Dr. Wilson Wang, IRC’s Senior Clinical Advisor, and his team to modify an electronic medical record system (EMR) that was originally developed for use at U.S. Veterans Administration hospitals, and to integrate it with a self-contained wireless network called the CliniPAK, which is designed to run off of any power source.

The computers—waterproof Sony Xperia Z2 tablets—are meant to be destroyed at the end of the unit’s life to protect patient data; the data itself would be anonymized before being distributed to other doctors on the network.


The original VA-based records software comes from Vecna’s corporate arm, while the hardware is built by Vecna Cares, its charity arm. “You can literally plug [CliniPAK] in to solar or battery and launch an [EMR] system with the touch of one button. You don’t need IT personnel, and it could run for days without maintenance,” says Theobald.

These features would be crucial in situations where infrastructure is lacking, and where imported technologies carry with them the shortcoming that, without the expertise to operate them, gadgets often go unused. Other CliniPAKs are deployed in Kenya, but this is will be software’s first run. (The IRC hopes to start treating patients with the new system in the middle of December.)

In the VA, the patient information exchange manages queues, orders, and data for thousands of users, and, not surprisingly, plays an important role in billing. But patients in Ebola clinics in west Africa are not billed, and (so far), they do not they file lawsuits for bad care.

Wang’s team has retooled it to suit their, and the patient’s, needs, streamlining the Ebola clinic in three ways. It manages staff shifts and patient information, it records clinical data and sends it to a cloud-based server for near real-time analysis anywhere, and it suggests treatments to health care workers.


“There was an opportunity to design electronic medical record [software] only for quality, to support the providers, and to make sure that the patients got the treatments that were needed,” Wang says. “We call our system JEDI [Joint Electronic health and Decision support Interface]. It was a joke, but we’re serious.”

For a doctor or nurse who spends up to 90 minutes in a hot plastic suit and is treating a patient who is seizing, for example, the system is meant to serve as a helpful assistant. It would reference the patient’s weight, advise phenobarbital at 20mg per kg, “and then give you the actual dose,” Wang says, “So there’s no guessing.” And every step is recorded.

All of this may seem amazing given that it is being deployed for Ebola, a disease that, until this year, was treated far from cities, in unpredictable intervals and using relatively basic materials. Pre-outbreak, it was not a pressing global concern. Ask any infectious disease expert why Ebola didn’t have a high-tech tool-kit like experimental treatments, vaccines, or even a data management system at the ready, and they will cite those obstacles.

But it is also late. At this point, the outbreak has been raging in three countries and threatening others for almost a year. It has surpassed the worst previous outbreak 40-fold; months have passed since the World Health Organization declared Ebola an international priority. The World Bank, the IMF, the U.S., and many countries have poured hundreds of millions of dollars into fighting it. And this is 2014, not 1976. Why is technology only catching up now?

Funding has helped. There are now enough funds to try throwing more than just tents and personal protective equipment at the virus. And many organizations are involved, some of whom are primed for innovation. For Debbie Theobald, using electronic medical records is an obvious step. “This is just best practice for how to set up a system,” she says.

Photo: via Twitter user Rachel Unkovic / IRC

Good Data As A Patient’s Right

Another big factor in the clinic’s upgrade: Responders are realizing the need for better treatment and data collection. Collecting data about how Ebola attacks a human body is notoriously difficult, for the simple reason that no material that enters a clinic can leave it, including paper.


“I’ve seen or heard stories of very creative ideas,” says Theobald. “They yell across the barriers to people, they fill out all the forms and post them in a window for people to copy. These are not optimal solutions.”

A study about this very problem was published in the journal Viruses in December 2013, just before the current outbreak exploded. The authors spoke with clinicians familiar with hemorrhagic fever clinics and asked about recording shortfalls. Despite more than 35 years of the disease, there were many. Low- and high-tech solutions were proposed, but it has taken months to implement the simplest of these in field clinics. Infrastructure, relationships with health ministries, and logistics remain real obstacles for Ebola responders. But, the study noted, new approaches could potentially do a lot of good.

The paper received mixed attention, in part because it was published when Ebola scientists and responders “go back to their day jobs,” said one health information manager. But Wang, of the IRC, was inspired by the study. It helped that his team already knew about Vecna’s software.

“Being able to enter [data] in the red zone and have it be available everywhere immediately, without the errors, is a huge operational and clinical benefit,” says Theobald.

The data, says Wang, will be given to the Liberian Ministry of Health to aid in tracing the spread of the outbreak, while some of the other data will be anonymized and shared with the rest of the world. Knowing what specific Ebola treatments are being provided when and where, how many patients are in the unit, what their survival rates are, and where survivors go after they leave the clinic can help doctors and researchers better understand the current outbreak and future ones too.

Leaving data in the clinics and treating patients in a rush costs lives, in West Africa as in America. Saving people means treating them early, measuring their treatments, and offering them novel approaches, like the kind every American and European who has contracted Ebola has received.


EMRs declutter the process for a health worker in a clinic, but they also provide a platform to study new therapies like antibodies, blood serums, and regular daily care. Every patient treated can improve the outcomes for the next, provided their data is recorded. In the end, collecting and using medical data responsibly is a patient’s right. That is why without the high-tech gear, clinicians sometimes shouted vital signs from inside the red zone.

Another promising sign for the future of the fight against disease: The gear is readily available. Over the last five years, billions of dollars have poured into high-tech firms locked in the accelerating race for e-health. Systems that streamline patient services became profitable thanks to subsidies in the American Recovery and Reinvestment Act in 2009, specifically HITECH, the health information technology part. The VA’s interest in digital records has helped companies like Vecna; and that software can now be adapted to other, much smaller uses.

“If you come in with a full-blown EMR, it’s terrible for training, setup, and maintenance,” says Theobald. “We focused on having it be a lightweight, patient-centered medical record.”

The Robot As “Learning Experience”

Other technologies that have not yet been immediately accepted in the U.S. have encountered a valuable testing opportunity in the fight against Ebola. VGo, which makes the telepresence robot, is one of them.

VGo’s robot was created in Nashua, New Hampshire, and not for health care, but for industrial applications and busy people who needed to be in two places at once; it’s been used by Audi and by Boeing, and was developed in part by Grinnell More, who used to run iRobot’s Military Systems Division.

VGo is essentially video chat on wheels. Some elite hospitals use it and similar telepresence robots for in-hospital and long-term home care, sending it home with kids so they can connect with the doctor without traveling back to the clinic (this is also meant to save the doctor’s time). Executives use it for remote appearances and consultants at Audi use it to deliver remote technical assistance.


Bern Terry, VGo’s vice president of sales, said his team started thinking about the robot’s potentials in an Ebola context when the virus reached Dallas, because—and this is crucial—VGo does not have to be touched to operate.

“Ebola has opened new opportunities, because people say, ‘wow, we could use this’,” he says. Selling for slightly less than $10,000, it’s an expensive device. But Terry justifies this against the cost of every suit that needn’t be thrown away just to enter a clinic. (Bloomberg puts a suit at $77, and there is also the danger avoided every time a suit is handled, which is harder to put a price tag on.)

Other robots have been proposed for fighting Ebola in West Africa, including one originally designed for space exploration, and Baxter, a workplace robot that could, some say, be used to help workers remove contaminated clothing. In October, the U.S. military said that “Saul,” its light-blasting disinfectant bot could also be used to clean up Ebola.

After Monrovia, Debbie Theobald will take her VGo bot into the red zone of an Ebola treatment unit in Sierra Leone, where it will help her train health workers to use the EMR she is installing there next. (The local Wi-Fi network extends just a few hundred feet. The connection to the rest of the world is not available yet, but when it is, a doctor in New York or Atlanta could connect directly to a patient in Africa, VGo says.)

This process isn’t easy. In the developing world and this part of Africa, new technology is treated with caution by the aid organizations that deploy it. Most foreign humanitarians try to anticipate how new technologies will be received and how robust they will continue to be once the team leaves. Unsurprisingly, Terry Bern is optimistic. “If it works, we’re likely to send a lot of those over there.”

Clearly, there is demand. The first VGo was sent only to help Theobald train staff to work the EMR software, not to see patients; it will travel with Theobald to assist with her next installation in Sierra Leone. The bot has piqued interest however: another VGo is being sent to the IRC clinic in Monrovia, with the aim to try it in a live patient context.


Theobald says using the robot would be a learning experience for Vecna when approaching various future robot-appropriate scenarios. “Originally it was for training purposes, to be with the doctors and nurses,” she said, “but I hope that people will feel more comfortable with this technology and think of other uses for it to proliferate.”

And if and when a robot does roll up to a sick patient, in a place with limited internet penetration, few smart devices, intermittent power connection, and no other robots to speak of, a place that has suffered terribly while the international community has mobilized, a place where clinics are alternately feared and respected, a place where health care workers have poured sweat and lost lives to serve; when the robot, gleaming of American ability and techno-optimism, broadcasts a human face to a sick person in a bed, that will be a very interesting first.

Colin Baker is a writer, radio and video producer based in west Africa. You may find more of his work at