The BRCK team sets up at the Rhino Charge 2014

We tend to be an adventurous bunch, and so it shouldn’t be surprising that Philip was the navigator for the BRCK vehicle in Kenya’s annual bushwhacking event called the Rhino Charge.

The Rhino Charge is unique, it works this way…

No one knows where the race will be until a week in advance, when a “jump off” town is named. People drive there the day before and are given a map to the camp. The night before the race, everyone is given 14 coordinates, with one as their starting point (each of the 65 cars are put at different ones, so about 5-6 at each node). The next morning at 5:30am they’re driven to their starting spot, and at 7:30am the whistle blows and they’re off. Each team’s job is to find the shortest route to all 14 spots and checkin within 10 hours. Whoever has the shortest distance between and finishes all 14 checkpoints, wins.

The BRCK Mobile – car 18

Rhino Charge vehicles checking in for scrutineering before the race

A couple of weeks ago the head of Rhino Charge got in touch with us to see if we’d mind setting up a few BRCK units for the competitors camp to allow people to get on WiFi there. We agreed, and soon enough Emmanuel, Jeff and myself were also going 6 hours north of Nairobi to Samburu country with Philip to setup some BRCKs.

Setting up the BRCK

It turned out that the Rhino Charge competitors camp and headquarters were about 10Km away from the nearest mobile phone tower. This meant that you could get a 3g signal, but barely, and it would fade in/out all the time. We were solving this problem at the competitors camp with the use of a directional antenna and mobile signal amp (booster).

We then got an urgent call from the head of Rhino Charge, HQ had just lost their internet and were wondering if we could switch our operations to their tent instead. Off we trundled with our gear and did the following:

The BRCK setup at the Rhino Charge HQ

• Directional antenna (by Wilson) pointed at the tower 10km away
• Wireless amp booster (by Wilson) to increase the signal strength
• 12v battery (by Brunton) to power the amp
• BRCK to transmit for WiFi
• BRCK with an ethernet cord to plug into a TP-Link 3420 to broadcast WiFi to their IT staff
• Safaricom SIM card with 32Gb of data
• Solar battery (by PowerMonkey Extreme) to charge the BRCK

Philip sets up the BRCK in the Rhino Charge HQ

Me setting up the directional antenna, using my car as a ladder

Takeaways

Whenever you do field testing, there’s always a sense of relief when things work as they should. In this case, it was extremely gratifying to be able to help the Rhino Charge HQ staff on their connectivity needs, and it validated a lot of what we say about connectivity when you’re in a place with poor infrastructure. The fact that we were able to get the BRCK up and going, provide the backhaul for all of the media and stuff happening around HQ was fantastic. The only time we lost signal was for a 5-hour spot when the main Safaricom tower we were connected to got hammered by too many spectators showing up and trying to check their Facebook feed… 🙂

This does go to show that when your last line of connectivity is the mobile provider, you’re really at their mercy for what is available on their mobile tower.

Moving the power button from the top to the side was a decision we made after our Turkana expedition. This was a good move, with the dust flaps on the side, we cracked open the BRCK when we got back and there was no dust inside the case at all.

When we were struggling to get connection, we needed to know that the BRCK needed to show us that it was struggling and what the issue was. Our redesign of the lights gave us this, and it was extremely helpful this time up north.

However, there’s also some lesson’s learned and things we want to improve on the BRCK, which is good news for everyone buying a device. When the BRCK team “eats our own dogfood“, you benefit. Here’s what we found that we didn’t like, and which we’re fixing (or have already fixed):

• We need to come up with a model that defines what “abnormal BRCK behavior” is. This would be extremely useful for sending out SMS alerts especially where the device is deployed in a remote location. To put this into perspective, Jeff and Emmanuel had to periodically drive to the HQ to check on the BRCK! (like a bunch of field officers in an NGO) [Working on, will be done before shipping]
• BRCK not turning off: On the day of the setup in the field, we attempted a reboot but the BRCK wouldn’t turn off. To circumvent this, we had to pop the top open and perform a short on the BRCK board. [Fixed]
• Spotty Connectivity: Before connecting the Wilson antenna, mobile network connectivity at Rhino Charge HQ – via the BRCK -was extremely spotty; signal strength was below 22dBm. We have no way of communicating this to the user. In our opinion, the lights aren’t sufficient to communicate this simply because it’s too much information for the user to memorize. Recommendation: Communicate this info via the on-board dashboard. [Working on, will be done before shipping]
• Detection of connected network interfaces: Both the 3G and Ethernet network interfaces were connected and active but this information was not been reflected on the cloud dashboard and preferences page. [Fixed]
• Local SIM metadata: The phone number of the SIM and it’s respective mobile network wasn’t available on the cloud (and is not yet being pushed to the cloud). Having this information would have made balance top up of the local SIM less of a pain. [Working on, will be done before shipping]

(Note: I’ll do another post later on the challenges, constraints and opportunities of designing in Africa, but here’s a real case study for why it’s important.)

Reg and Philip spent a day last week at Amboseli working with a group of conservationist called Lion Guardians, who have a permanent camp about 10km from the edge of the Amboseli National Park. They went at their request to see if BRCK could solve their current connectivity challenges before they invested in satellite equipment.

A Problem of Too Many GSM Towers

The situation is that Lion Guardians can get decent signal from one operator (Airtel) but very slow data connections. They can also receive signal from another mobile operator (Safaricom) in certain physical locations within the camp but it requires standing in one spot and holding their phone in a certain way. One of the workers is able to send/receive data by holding his laptop in one of these spots to get a decent data connection.

Reg and Philip spent a couple of hours surveying the signals with a combination of the OpenSignal Android phone app and the Wilson boost equipment. They found a very strong Airtel signal but it only provided edge data. They found more than one Safaricom signals, one of them provided no data – even with strong signal – and another provided reasonable edge performance. However, when locating in the exact spot identified, they were able to get 3G signal and faster data.

After attempting to secure a reliable signal from the identified 3G tower (they were tracking tower IDs) they stumbled upon a hypothesis for this type of situation, which isn’t an uncommon issue.

The hypothesis is that the 3G tower is the tower that is furthest from camp and therefore has the weakest signal. From most locations in camp, one of the stronger signals from the other towers is given preference and no/poor data connections are the result. However, when standing in a location that blocks the stronger towers (due to a building and a tree), the weaker signal take precedent and a solid 3G connection is possible. As we understand it, the device will usually “hear” several towers and then has to make a decision about which one it’s actually going to talk to. Signal strength, not quality, tends to be the easiest metric to engineer for.

A BRCK GSM Lock-on Solution

The opportunity is to create a means of forcing the selection of a specific tower ID even if other towers have stronger signals. In order to achieve this improved result we are going to need to address two principal issues:

• The first is that we have to develop a means of indicating to the BRCK which tower IDs are acceptable and which should be rejected.
• The second is to develop a means of doing effective site surveys that will allow anyone to identify the best carrier/tower combination for a given location.

If we can do this then we should be able to offer a very competitive performance improvement in locations with mixed/limited coverage. You can imagine being out with your BRCK in a new environment and getting a list of possible connections, and choosing the best one – not the strongest one, in the future.

This observation further reinforces our “engineered in Africa” position since we would likely not encounter this scenario in the West. The fact that Safaricom has multiple towers but only some of them have good data connections is a very unique problem for the emerging markets.

With respect to Lion Guardians, they have been told that a new Safaricom tower is coming that will be close to camp. This will likely provide them with the performance they are looking for. However, they will still need a BRCK since they want to have a single, reliable, self-powered connection that is shared using an outdoor Ubiquity WiFi AP.

We started out early from Kurungu towards Baragoi, we didn’t have any issues with bandits as there was an increased police presence in the area. It had rained the night before, so the ground for our whole trip was a bit damp, perfect for fast riding. It also created some very waterlogged areas, where we had to get completely soaked up to our waists in the water as we rode through.

Taylor Martyn riding through a deep puddle, north of Maralal, Kenya from WhiteAfrican on Vimeo.

Until we got to the rocks near Maralal, it was a fairly easy ride, then things began to toughen up. One puncture on Taylor’s bike, a quick change, and then off on our way again put us into a rainy Maralal by lunchtime.

We went through some massive puddles along the way, so our boots were completely soaked and we were then able to empty the water from them as we enjoyed some hot chai, chapatis and stew at a local hoteli.

Back into our wet socks and waterlogged boots, and we were off for some rocky, wet and slippery riding down towards Rumuruti – where the tarmac roads begin.

From there it was a quick run down to Gilgil, which we reached by 6:30pm and started for Nairobi, only 1:30 ours away (or so we thought…).

Now, due to the number of spills the bike that I was riding had endured, none of the lights worked, except for the headlight and the right blinker. We put Taylor’s fully functioning bike in front, me in the middle, and the Land Rover behind. The car alternator had gone out the night before, so we didn’t run the car with much electronics all day and the battery seemed to be holding up well.

20 minutes later, just past Naivasha, my headlight went out.

Now, the only place more dangerous than driving on a road in Kenya at night, is the side of the road. But, there we merrily sat, Philip and Reg testing out the electronics and finally realizing that the bulb itself was burnt out. I wondered what a flashlight would look like inside the headlight casing. It turned out to be surprisingly bright, so Fady donated his super expensive and very bright bicycling flashlight to the cause, which turned out to fit perfectly.

As we duct taped that jua kali light into place, the Land Rover’s batter gave out it’s final death call and the lights went out.

No problem, we put it in reverse and pushed it back down the hill to jump start it. Fortunately, we had a 12,000 mAh Brunton Solo 15 solar charged battery on hand (similar to their new Impel), and we hooked that up to the vehicles lights.

Off we rode into the night, now the truck in front, me on my raggedy old cludged together light bike in the middle, and Taylor riding anchor. It was now very cold, as it gets in the Kenyan highlands during the rains. Then, as we got up towards the top of the escarpment, at the 8,000 feet high level, it started to seriously rain. Now, it’s already hard to see at night on a bike with a headlight pointing up and blinding pilots, with people’s brights deflecting off of the droplets on your visor it becomes even harder. Pop up the visor and you can see again, but the cold and stinging rain now hits you in the face. Better to live with a little pain than die in a collision, so up it went.

We had just passed the Kijabe turnoff when the Land Rover’s second battery died. Fortunately, we were in my territory, as my alma mater RVA was 15 minutes down in Kijabe. We pulled over at a petrol station, called up my old dorm parent Mark Kinzer, and he found a friend who was willing to donate us a fully charged 12v battery for the remainder of the trip. 20 minutes later we had a battery and were moving down the road again.

16 hours after leaving Kurungu, wet, cold and utterly fatigued, we arrived home and well in Nairobi.

Thank god for old friends, good (enough) roads, no bandits and safety through the night drive. A bunch of creative problem solving was needed to get through, as well as stamina and perseverance in the face of the friction that is thrown at you on a trip like this. I can’t tell you how happy I was to get some warm food at home, a warm shower to thaw out my body, and a warm bed to crash collapse into as I went comatose for the night.

Would I do this trip again? Absolutely. In fact, “BRCK Expeditions” are a new thing for us that we’ll continue to do as a company and that we’ll look forward to hearing about from our BRCK users around the world.

I wrote up another blog post on my personal blog today about some of the things I learned about our team. They are:

1. Work together, help each other selflessly.
2. Think creatively and push harder to get through the friction.
3. Always get back up, persevere.

There and Back from BRCK on Vimeo.

A quick sneak peak at our adventures up country thus far.

Yesterday we took off from Sibiloi towards Loiyangalani and then on towards South Horr. Besides a tire puncture along the way for the car, it was nice and easy. We were able to stop and talk about the eclipse with a few eclipse chasers from Spain and they said they got a somewhat clear picture of it, with just a small bit of cloud cover. There were so many vehicles leaving the north, that big clouds of dust covered us and made us look like brown powdered humans.

While stopped on the side of the road, about 100 Kilometers from Loiyangalani we could get a very weak Edge mobile phone signal. Enough to send some tweets and call people, but no real media uploads. We spent some time with the Wilson amplification antennas, but didn’t have too much luck because we didn’t know the exact location of the tower from where we were. It’s a great piece of equipment though, one we love to break out and use with the Yaggi antenna from the top of the Land Rover.

Fady got his second motorcycle riding lesson. Again, on some of the worst roads that Africa can throw at you… amazing really. He did it all with speed and grace, and just a single tip-over.

We overnighted at the Swanepoels again, where we talked about the route possibilities. Massive rains have cut off a lot of the roads. Our 3 choices were: Marsabit (safe, long), Korr (not possible due to rains) and Baragoi (security issues, but short). We were heavily cautioned about Baragoi and Maralal, so opted for Marsabit and set off at 5:30am.

Today’s trip started off with eating a lot of flying ants. They flew so thick as we rode that it was like riding in a snow storm, except with wings. I got the all in my goggles and helmet, light and other gear. Everyone was covered in flying ants when we stopped 30 minutes later.

2 hours later found us sitting by the side of the road wondering how we were going to unstick a very stuck Land Rover. We had slipped and slid our way through, and Fady earned his “wet” riding badge with honors.

4-hours later, we were still stuck, after trying every trick we knew. I even tried something very stupid, and paid for it, trying to pull with my Suzuki DR650 and burning out the clutch plate. Now my motorcycle isn’t running and I have to leave it in South Horr with my friend to get it back down to Nairobi in a few weeks. This, I’m not happy about, but I should have thought through what I was trying to do.

A Landcruiser came through in a few hours and helped pull us out. 2 riders and 2 motorcycles sat by the side of the road while the others tried to push through. However, a new lake had formed on the way to Kargi and Marsabit, so they turned around and came back. A Land Rover had showed up, and we hired him to take my motorcycle back to South Horr.

So, now we sit in the same place that we sat yesterday at the same time.

Tomorrow we shoot off towards Baragoi. Though there have been bandits on the road, we’ve heard from another friend who went that way today that there is a lot of extra security on it and that they’ve just graded the road. With a lot of luck, we can make it to Naivasha and onto Nairobi by tomorrow night.

Well, we had a great trek up to northern Kenya, to the shores of Lake Turkana to catch this rare hybrid eclipse (see past posts). However, we had a massive dust and rain storm hit 45 minutes before the event. Unfortunately, those clouds hit the sun just the few minutes before the eclipse.

All of us in Sibiloi saw the light dim out, but didn’t get to see the moon cross over the sun.

An airplane just landed, that we think got the eclipse from above the clouds. Wish we had one of those!

We’re now catching a partial bit of the eclipse as the clouds break open. Image shortly if we can capture it.

6am rolled around and we all jumped up, hungry, but rested. Sunrise was amazing. A beautiful color that shed some light on where we actually were. As I type this, there is a warthog about 20m away to my left, while both greater and lesser flamingos, along with some kind of wild goose, feed 20m away in front in Lake Turkana.

Sibiloi is distant and harsh. It’s a forgotten mixture of austere beauty, heat, wind and silence – that though it isn’t welcoming, is still a joy to be at.

I can think of no reason why 200-300 people would come here, except for the rare hybrid eclipse that we’re about to see tonight.

We swore we had driven through trees, but there were none around, not within 10 miles. This should be an object lesson in the illusions that one sees when riding and driving after so many hours on the road at night. It also became apparent that we’d have to lose any sense of privacy when it came time to find a restroom. It’s a good thing that this is not a mixed group. 🙂

All morning we’ve been working on getting a couple things tested:

1. Internet, we’ve got a BGAN terminal for satellite data connectivity provided to us by Indigo Telecom in Nairobi. They rent them out to people everywhere in the region, and have been great to work with. That’s working nicely, and hopefully we’ll get the eclipse livestreamed at brk.com/eclipse.

2. Mobile phone signal, Wilson has provided us with some long range amplification antennas. We tested them in Kurungu, where we turned a weak Edge signal from Safaricom that couldn’t transmit any data into a very strong Edge connection that was usable. They gave us a couple to test out, and we’ll do so again on the road back to Loiyangalane, from the a hill about 20-30km away. We don’t have a good line of sight from where we are to the tower in Loiyangalane from Sibiloi, so can’t get a signal here.

3. Camera setup for the Eclipse. What we’ll be streaming with is a Frankenstein-like beast, that’s cobbled together from a web cam, toilet paper roll and a Schneider large format lens, all held together by gaffer tape. It’s really amazing, the fact that the team has such a mix of camera and digital know-how to pull this off by Jon and Philip, was great to watch.

4. Power for the computers and BGAN. There’s a lot of “hack it together” knowledge on the team, and between Reg and Philip they were able to jimmy rig some solar to car battery to inverter to hacked together power plug. some people will tell you that this is “real field constructed power distribution” – we all know it’s a jua kali rig job.

While all that is being set up, it’s great to see the BRCKs humming along quite nicely. The BRCK flag now waves above the Land Rover at Camp Jawbone (named after the zebra jawbone we found when setting up camp last night).

Now we’re waiting for 5pm to get the livestream rolling. Crossing our fingers now for a cloudless sky tonight.

Update: It’s 5 minutes to the eclipse, we just had a massive sand storm and rain storm pass through. barely getting connection and a lot of clouds. Probably can’t stream, so will try to upload images.

Day 3 started a little late as we had to fix Taylor’s front forks, which had met with an untimely collision with some mud and sand the day before. He rode about 100 kilometers on a skewed front tire, which just made the sand at night a bit harder, so it was a miracle that he didn’t go over again.

After an amazing breakfast from the Swanepoels, we made great time on smooth roads from Kurungu to Loiyangalane, this day resolved to stop and eat lunch, refill the water and relax for 30 minutes before we did the last run to Sibiloi. Now, the road to Sibiloi has always been an adventure, and it didn’t disappoint us… It’s about 125km from Loiyangalane to Sibiloi, and it’s full of massive boulders, though those had been cleared and a beautiful road put up the incline of the escarpment.

Again, we made good time for the first half of it. Then, Philip took a bit of a tumble on the open sand flats. There’s a real tricky bit of riding where the sand becomes talcum under you with little warning. He hit a patch and as he went down hit his hand on a rock. A bit of blunt trauma treatment later it was decided that he couldn’t ride any longer. This left us in a bit of a bind as no others in the vehicle knew how to ride a motorcycle.

We were about to hoise the large Honda 650 XLR onto a passing Kenya Wildlife Services vehicle. Right before we lifted it, Fady decided he’d give it a try…

Fady is some kind of superman. He learned to ride a motorcycle on the sand flats at 5:30pm and rode for the next 3 hours into the night through the African bush. It was quite amazing really, I’ve never witnessed anything like this before, and know that I could never have done the same. Your first 3 hours EVER of riding a motorcycle are done in some of the harshest terrain – at NIGHT!?

Needless to say, we’ll be providing Fady with a certificate of excellence. (special note from Fady: “Smurfy don’t kill me”)

We were the final vehicles into Sibiloi National Park last night. It was about 8:30pm and we were completely exhausted. Now, you’re normally not allowed to bring motorcycles into any national park in Kenya. However, this time the warden made an exception, as long we agreed to ride straight to the camp and not ride them again until we left. I can’t tell you how good that made all of us feel and we owe the KWS a huge note of thanks for taking so much good care of us last night. In particular, a gentleman by the name of Sode was exceptional and I just wish all government employees were as pleasant and helpful.

Once past the gate, we drove down a road until we literally drove into Lake Turkana. After we backed out of the water, the KWS guys showed up again, along with Andrew Nightengale – someone I know from staying at his farm with my family a few years ago. He pointed us towards a place right along the lake a few hundred meters away.

Once again, exhausted at the end of another African bush night drive, after a full day on the road, we set up our tents. A few bowls of my wife’s chili later (promptly announced as the “best food ever” by everyone), with some Dickel Bourbon whiskey as a chaser, we collapsed into our tents and slept like the dead.

Well, yesterday was exhausting. 14-hours of driving and riding later and we pulled into Kurungu.

We roared off at 6:30am from Nanyuki, only to have a puncture an hour later, that cost us almost 1.5 hours due to complications. This had a ripple effect that meant getting to Laisamis at 1pm. Rain hit us in Laisamis, and softened up the soil on our way to Ngurunit. That slowed us down a great deal.

We had about an hour left of daylight by the time we hit Ngurunit, some of the most amazing scenery you’ll see in Kenya, but we couldn’t stop for pictures as we had to use all the daylight possible for driving.

Night driving is hard anywhere, but it’s harder when you’re in mud, deep talcum-powder sand and rocks. We had all of that and we did it for 3 hours in the dark. There were a couple layovers, but thankfully we were going at a slow enough speed that no one was injured – and no bikes were put in too bad of a condition.

There were times where we got so fatigued, and then were staring at that single beam of light on a bumpy dirt road, so that the hard angles all looked like a gully or hole. Made for slow, stressful and hard riding.

Here’s where I get to say thank God that I grew up here in Kenya since I was a little boy, with the Swanepoel family (originally from South Africa). Father and son live up north here, and they welcomed us late at night with a warm meal and a roof and beds.

This morning we woke up to hot coffee, 800mg of Ibuprofen, bruises and soar arms/shoulders. Yesterday was hard. Today is harder. We’ve got a fairly easy go of it to the lake, then it gets hot and hard as we ride alongside the lake to Loiyangalane. After that, it’s even worse, so we here, with boulders. We’re about 30 minutes behind schedule, but I think we’ll make it okay, and hopefully in the daylight.

This morning finds the BRCK team in Nanyuki, only about 3-4 hours outside of Nairobi. The same place we were supposed to be having breakfast yesterday, not today. A rather inauspicious start happened about 30 kilometers into the trip when the Land Rover decided it was going to have some cooling problems.

After two hours of working on solutions, we realized that this Land Rover just wasn’t going to make it. Too bad, as it had two tanks and better range than most. Fortunately, Fady Rostom (of Ark/Bonk) is traveling with us. He offered up his Land Rover 110, and we quickly got in touch with a mechanic in Nairobi who put a roof rack on his vehicle in about 30 minutes.

1pm found us with all equipment unloaded near Thika town, on the side of the road. 5 tires were changed from one Land Rover to the other, and we repacked the new vehicle. Finally, we were on our way (again)!

In hindsight, what was extremely disappointing at the time ended up happening about the best time we’d like to have seen a problem like that develop – in the beginning of the trip near the big city, where we could still make a change.

We got into Nanyuki in the evening, after a nice cold rain hit us, much to the chagrin of those of us on motorcycles with only mesh jackets on. Taylor had his full rain gear (who we had been making fun of earlier due to the heat), and while we froze he rode in relative comfort. Dinner was at Barnies, and then a shack that makes and sells the best cheeses that we’ve ever found in Kenya, called Silent Valley Cheese. 6 kilos of cheese later we found a house owned by a friend to sleep at for the night.

This changes the stages of our trip, and now we’re very glad that we built in a buffer day for the drive up to Lake Turkana. We’ll still arrive the day before the eclipse (Oct 2), but in the evening instead of the morning. Our route now entails an early morning offing from Nanyuki, final fill up of fuel in Isiolo, then off road around Laisamis towards Ngurunit and on to Kurungu and Soth Horr for the night.

This next section along the edge of the Matthews mountain range is one of the most scenic places in Kenya. We’ll stop along the way for some pictures, and then get into Kurungu in time to test the Wilson amplification antennas. The point of these amp antennas is to extend the range one can be from the mobile phone tower to get internet connectivity. We’ve used one before in Lamu, now we’re trying it in the bush, and with the BRCK.