The Unique Challenges of Making Technology Work in Developing Countries
Making technology work in the rural and developing world is a process full of unique challenges. I am Andris Bjornson, Inveneo’s Chief Technology Officer and practicing project engineer. As an Inveneo engineer, I have to be prepared for it all. Recently in Kenya’s Masai Mara Game Preserve, I had what I’d call in rock climbing terms an “epic” day traveling to bring the Cisco Community Knowledge Center (CKC) in Sekenani, Kenya (photos) online with high speed terrestrial broadband Internet.
This is not by far the most difficult deployment (that would be 3 weeks in Nepal dealing with monsoon rains and leeches) nor the most chaotic (probably our Haiti Earthquake disaster response) However, the condensed timeline, wild game, and bad weather combined to make this a good story.
I had arrived in Nairobi after finishing a week-long wireless connectivity assessment in the UNHCR Dadaab refugee camp and I was asked to tack on an additional day of work for this special project, to make the final adjustments to bring the connection live. This was the first time in Kenya that we would be connecting one of the CKC’s with broadband Internet by partnering with a local carrier at their tower and then using long distance wireless (WiFi) to connect to the CKC facility 20 kilometers away. It was supposed to be a quick in-and-out flight into the Masai Mara Game Preserve, as the CKC is just inside one of the Park’s eastern gates.
We usually send one of our Inveneo Certified IT Partners (ICIPs) to complete a technical installation like this. All of the equipment had already been put in place, but the connection had to be tested and completed now that the CKC had signed the service contract. First-time installations are always tricky.
Here’s how it actually went:
6:00am – Wake up in my hotel in Nairobi, scarf down a quick breakfast and call Bernard, a friendly taxi driver I had met earlier for a ride to Nairobi Wilson airport. This small airport is Nairobi’s jumping off point for all domestic, tiny aircraft that would generally worry my mother.
8:15am – Arrive at Wilson after navigating Nairobi’s legendary traffic and check in at the last possible minute.
9:15am – Take off in a Twin Otter: the favorite 20 seat, twin-prop workhorse of rural airlines from Nepal to Borneo.
10:30am – Land at the dirt airstrip at Keekorok in Masai Mara. Connect with Tony, the friendly Land Cruiser driver and guide whom I’ve hired to take me where I need to go. The change from UN convoys with armed escorts in Dadaab to convoys of open top matatus (vans) full of tourists seeking sightings of the “big five” (elephant, lion, leopard, black rhino, and water buffalo) is jarring.
10:50am – While bouncing down the dirt track to my destination, I see a herd of elephants! (Never seen them in the wild before.) Check in with the cellular engineers I’m scheduled to meet. Our wireless equipment, which serves the CKC with Internet, is co-located at their Talek Market cellular tower, and I’m supposed to meet them at the project site.
I learn that, though I’d made them promise to meet me at their tower by noon when I spoke to them last night, noon had become 1pm by the time I talked to them from Nairobi in the morning, and now it’s going to be 2pm due to car trouble. This is not uncommon, though given that my return flight leaves the airstrip at 4:15pm and the airstrip is at least 30 minutes from the tower, it is less than ideal. I need to connect with an international flight the same day in Nairobi, and this is the last flight out of Masai Mara for the day. I hope furiously that an hour and a half onsite will be enough.
11:10am – Arrive at the Sekanani CKC,the site I’ve come to connect. I meet with Kerry and Musa, the CKC managers, and discuss the problems. Initial diagnosis shows that what we’d thought was the cause of their Internet troubles is actually several completely unrelated problems. The Inveneo R3 server’s BIOS (the most basic “brains” of a computer) has reset itself and lost its settings, which can sometimes happen due to unusual power conditions. I fix this and show Kerry and Musa how to fix it if it happens again.
One of the server’s two mirrored hard drives has also failed (not uncommon in hot, humid and dusty environments. Sekanani is all three). Fortunately, our Inveneo R3 server has RAID: a technology that allows the server to keep running even with one failed drive.
12:30pm – Learn from Tony the driver that the normal road from Sekanani to Talek Market (the cell tower location) is deep under water from recent heavy rains. This turns a 15-minute drive into a 40-minute one. I pass this information on to the cellular engineers, who are still en route.
1:00pm – Quick lunch, then head toward the cell tower with Tony.
1:50pm – Herds of Antelope cross the road in front of us as we drive.
2:10pm – We arrive at the Talek Market tower, and find that the cellular engineers have just arrived! I eye our equipment at the very top of the 40m tower (a Ubiquiti NanoBridgeM5-25) and think how lucky our engineer, Jen, was to have installed it during the great migration season. She has pictures of massive herds of animals passing under her.
2:15pm – I discover that, three weeks earlier, the cellular provider moved our equipment and wiring to a completely new cabinet without letting us know. The way the move was done seems to have caused a power spike that wiped our long- distance WiFi radio’s settings. It’s a problem we knew about, but one we discovered after the equipment installation had been performed three months prior.
3:15pm – I reconfigure the radio from scratch, grateful I don’t need to climb the tower, and bring the link between the tower and the CKC back up and functioning again! Check in with the cellular providers’ NOC (network operations center) to verify connectivity. They confirm the test passes! This is good, because it has started raining and I’m running out of time. Upgrade firmware: the link to the CKC goes down! I’m out of time, it’s continuing to rain, and given the results of the NOC’s test, I theorize that I can fix the remaining issue remotely from Nairobi.
3:30pm – Dash for the airstrip
3:35pm – See some zebra hanging out by the side of the road. They don’t seem to mind the rain.
3:40pm – I receive a call from the NOC that their test has, in fact, failed. This problem will make it impossible to access the equipment remotely. Wish they would have told me this before I left the site, but I’m impressed with their follow-through. Better to find out now than after leaving the country.
3:45pm – We make the judgment call to push my return flight to Nairobi and my onward flight to San Francisco back by a day to finish the work. Spend 15 minutes on hold with the travel agent’s emergency line in an effort to get them to change the international flight. The international call burns through my remaining cellular credit and I never do reach them. Tony is also out of phone credits so I can’t even call Air Kenya to cancel my domestic flight.
4:05pm – Continue to the airstrip to tell the Air Kenya pilot that I can’t take his flight back to Nairobi. Buy cellular credit from a driver at the airstrip, call Air Kenya, and learn that the 4:15pm flight tomorrow is full. The best they can do is 10:15am tomorrow. Hope that buys me enough time to finish the work. Head back to the CKC to see what I can do from there because it’s much closer than the tower.
4:25pm – Try again to change my international reservation. I find the number for Swiss International Airlines. I ask the driver to stop the car so that the phone agent can understand me. Just as she is explaining that she can’t change the reservation because it was a Swiss Air flight booked as a Lufthansa codeshare, a giraffe casually wanders up and stops about eight feet from the car. I swear he’s looking right at me saying: “Clearly you should have called Lufthansa.”
4:35pm – We reach the CKC. I tweak some WiFi link settings. This fixes part of the remaining problem but not all of it. Then I have the “a-ha” moment realizing what happened during the firmware upgrade at the tower and that I need to go back.
5:15pm – Head back to the cellular tower (again, the long way because the fast road is washed out). It’s raining really hard now.
5:55pm – Exit the Masai Mara Park gates (the tower is just outside the park) and beg the ranger to let us back in in 20 minutes (15 minutes after closing time). Hope that I can fix the problem in 20 minutes. Pull on my raincoat, and grab my laptop.
6:05pm – Find myself hunched over my laptop, plugged into equipment inside a cellular equipment cabinet with two guards and a driver holding golf umbrellas over me to avoid drenching at least a hundred thousand dollars worth of cellular equipment and bringing down cellular service in the immediate area (our WiFi gear is housed in the same cabinet). Laugh at the absurdity of the situation.
6:10pm – Fix the problem while keeping surprisingly dry. I’m annoyed at myself that it was partly a stupid mistake on my part and partly a known bug I should have remembered, that causes a critical checkbox to check itself when firmware is upgraded to this particular version. Make a mental note to tell our teams in Haiti about this bug.
6:15pm – Call Kerry at the CKC. Still under the golf umbrellas in the deluge (though thankfully we’ve closed the cellular cabinet), talk him through some technical troubleshooting steps to verify that everything is working from his end. Hooray!
6:16pm – We all have to walk through ankle-deep water to get back to the car. The land around the tower has turned into a lake. Feet are soaked and muddy all around. Pass the tower guards some Kenyan Shillings for going way above and beyond the call of duty and helping me (and much more importantly, the equipment) keep out of the rain.
6:20pm – Start the trip back to the CKC, grateful that the Masai Mara ranger honors his promise to let us through the gate slightly after hours.
6:45pm – Spot some ostriches: cool!
6:50pm – Glad to reach my colleague who is passing through Nairobi for some meetings. He has good Skype access and makes the international call to change my flight. Change fees are totally reasonable! Phew.
7:00pm – Glad that Tony is a flexible guy, arrive at the CKC. This is way more driving than he bargained for. Do some general IT housekeeping, verify that the link is performing well.
7:20pm – It’s well after dark now. Needing to use the restroom, I stumble off with my flashlight to find the outhouse. Kerry warns me to watch out for hyenas. He is not joking.
7:30pm – Shake hands with Kerry and Musa, glad that everyone’s hard work paid off and that the CKC now has a 2Mbps symmetric Internet connection. This might not sound like much in San Francisco… but this is by far the fastest, most affordable Internet available out here in the bush.
8:00pm – Return to the safari camp where I had lunch and explain that, yes, I was supposed to have left already but please do they have room for me. Fortunately, they do.
8:15pm – Grateful to change into some dry clothes. Finally unwrap the soaked ace bandage from my swollen ankle (I sprained it a week before jumping on a trampoline with my wife). Relaxing at this amazing spot is a fantastic reward for an exhausting day of work. It all feels worthwhile because the CKC has Internet now and can greatly expand their (ICT) services to the community.