Developing a GIS-based roads maintenance management system
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 295 | |
| Author | ||
| Contributors | ||
| License | CC Attribution 3.0 Germany: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/43314 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
| |
| Keywords |
FOSS4G Bucharest 2019269 / 295
15
20
28
32
37
38
39
40
41
42
43
44
46
48
52
54
57
69
72
75
83
85
87
88
101
103
105
106
108
111
114
119
122
123
126
129
130
131
132
137
139
140
141
142
143
144
147
148
149
155
157
159
163
166
170
171
179
189
191
192
193
194
195
196
197
202
207
212
213
214
215
216
231
235
251
252
263
287
00:00
Data managementSoftware maintenanceComplex (psychology)Computer networkFrequencyUsabilityPhysical systemInterface (computing)Decision theoryPresentation of a groupSoftware maintenanceUser interfaceSoftware crackingMetreArithmetic meanAreaQuicksortProjective planeContext awarenessEnterprise architectureInterface (computing)Software developerDecision theoryData managementDialectReduction of orderMaizeNetwork topologyIdentity managementSoftwareComplex (psychology)Group actionCAN busHand fanMultiplication signReading (process)Computer animation
03:37
Component-based software engineeringSoftwareServer (computing)Process (computing)Projective planeSoftware developerInterface (computing)Cartesian coordinate systemPhysical systemCausalityComputer architectureTable (information)MappingWeb 2.0Dressing (medical)Front and back endsMultiplication signLevel (video gaming)Product (business)Server (computing)Quicksort
04:27
Physical systemOpen setOpen sourceComputer configurationSoftwareCondition numberTwitterSoftware maintenanceComponent-based software engineeringPhysical systemExterior algebraArithmetic meanGoodness of fitRoutingSheaf (mathematics)Office suiteFunctional (mathematics)Service-oriented architectureVotingLatent heatLine (geometry)TwitterSoftware maintenanceQuicksortTrailMultiplication signBasis <Mathematik>Self-organizationConnected spaceCalculationUniform resource locatorNumberWordInternetworkingCentralizer and normalizerMereologyCondition numberMathematical analysisSoftwareInternet service providerSet (mathematics)Level (video gaming)Limit (category theory)InformationComponent-based software engineeringGroup actionComputer configurationExpert systemUser interfaceView (database)CASE <Informatik>Open setTable (information)Open sourceCoordinate systemPoint (geometry)Interface (computing)Computer animation
09:51
Software maintenanceTwitterAlgebraic closureUniform resource locatorNumberInterface (computing)Trail
10:36
Mathematical analysisFormal languageTwitterProcess (computing)Open sourceNatural languageIncidence algebraProcess (computing)TwitterCondition numberShared memoryMathematical analysisElectronic mailing listMaxima and minimaComputer fileSheaf (mathematics)Level (video gaming)State of matterCarry (arithmetic)Metropolitan area networkView (database)SoftwareMedical imagingInterface (computing)Component-based software engineeringCausalitySubject indexing
12:28
SmartphoneSelf-organizationLevel (video gaming)Open sourceFront and back endsOpen setSet (mathematics)Term (mathematics)WhiteboardReading (process)Process (computing)Mathematical analysisInterface (computing)TrailView (database)Point (geometry)TwitterMobile WebCartesian coordinate systemInformationBasis <Mathematik>BitOrder of magnitudeCausalityNatural numberTime zoneMobile appThomas BayesComputer animation
Transcript: English(auto-generated)
00:12
of developing a GIS vehicle maintenance management system. She comes from Kenya, and let's start.
00:24
Okay, I'm Laura, as she said, I'm from Nairobi, Kenya. I currently work as a GIS analyst developer for a social enterprise called Sanadu, which is in the sanitation sector, but this is what I did for my undergraduate research
00:43
earlier this year. Yeah, so maybe just some context on why do a project that is focused on roads? In some countries, especially in Africa, like in Kenya, roads are the main form of transport
01:02
from a town to a town, a city to a city, and even for some regions like Eastern Africa, from country to country. So basically roads affect all the activities that one gets to do from morning to evening, night to morning, yeah. So as I've said, it affects very simple activities
01:22
like going to work or a shop, or very urgent and complexions like firefighters or an ER team getting to an accident or an incident. However, just like any other infrastructure, roads deteriorate with time, and therefore there's a need for management.
01:44
And road management sort of includes both the development of new roads and the maintenance of existing roads, where the development means basically increasing the network, because also for most developing countries, not all areas are covered by this road network,
02:01
if we are to say tarmac roads. And also the maintenance of the roads that are existing so that they can be usable for a much longer period. And so basically also when you go to a developing country
02:21
or a third world country, what you'll notice is a lot of funds are focused on creating new roads, and roads are entirely not maintained at all, meaning you're creating a new road and I can walk a few meters away and find a road that's full of potholes and cracks and not usable. So basically the idea was to create
02:42
a very simple web interface that will aid the concerned road agencies in Kenya to be able to make better informed decisions as road maintenance is concerned. And by creating this interface or dashboard, you'll be able to provide evidence on how public funds are managed,
03:01
and if you're not aware, Kenya is one of the countries where employees actually give one of the most highest amount of taxes around the world. And most of these funds are usually misused, so this interface would provide a way
03:20
to show how funds are used and sort of reduce corruption and improve all other sectors that are affected by road transport and reduce other road related accidents that claim hundreds of lives per year. So basically, to create the interface,
03:44
it's as simple and as common as it is, you'll need a web server and use them since it's still locally hosted and the product is in development. Then a map server, use your server, then Azure database, use Postgres.
04:00
And then for data preparation, use QGIS and Mapbox Studio for the web maps on the application. And then just for the front end, you just ReactJS just because at that time everyone was using React and Leaflet and Python for the backend. Yeah, and this is sort of the system architecture
04:22
of how the whole application works. So yeah, just getting back to like what was mentioned yesterday in the keynote was you might have software, you might have the skills, but if you do not have the data, then the software is sort of useless.
04:42
And that was actually the main challenge when coming up with this solution is that roads are being maintained in a sort of haphazard way. Like there's no means that they'd say they're selecting a road just because they didn't have data.
05:01
And the data that was available had been collected decades back using an initial navigation system that was like put in a vehicle and then the vehicle is supposed to go through all the roads and calculate the acceleration and how it moves so that if it moves down, then they can say it's a pothole
05:21
or it's a crack or something. But then that was just too expensive to collect data and you're not even thinking about the maintenance and all that, yeah. So they sort of stopped doing it, they did it once. And then the government also explored other options like using Lidar data and UAVs
05:41
which they also claimed to be too expensive for just data collection. So for us to have developed a solution that will then be actually used by these agencies, we needed to look for open data sources and crowdsourcing options that will actually be low cost or no cost at all.
06:02
That's when now we got to use OpenStreetMap data and transit data that had been collected by an organization called Digital Matatus in Kenya. Then also they recently released Uber Movement data last year and also this year when Nairobi was also included.
06:21
And then Twitter data. And then now data that was provided by the government was on maintenance history and drainage data. So maybe just a few things was that OSM was the best data and then maintenance history was basically data on which roads have been maintained in a year
06:41
and which roads have basically not been maintained since they were constructed. And then transit data basically showed the relationships between our users and roads and how they use roads on a daily basis. And then Uber Movement data also had data on traffic
07:01
and speed and all that, yeah. And then for Twitter data now that's where the crowdsourcing came in. So basically Kenyans are active on Twitter and we noticed that on a daily basis from morning as early as 6 a.m. people are tweeting about road conditions, if it's an accident, if there's traffic,
07:22
if there's a pothole, if there's something, something on the road that happened to them. So basically we used a trippy API which basically searches data from Twitter that has been sent with the location component before Twitter disabled it.
07:42
Yeah, so basically we got a tweet, all the information, whether it's the name of the road and actually the GPS coordinates if that person shared those coordinates. And then now using the software and the data then developed the web interface.
08:05
So basically it's just a map with the roads and analysis of all the data sets combined to provide roads graded from A to E where E was like the best condition
08:22
and E was the worst condition. And then since most of these road agencies, the people who are working there are not GIS experts, we also provided a table view just in case someone might have issues or might just want to know the specific roads then assign the funds.
08:42
So we also provided a table view. So also another fact about Kenya is that we have a developed government. So we have 47 counties which manage funds per county. It's not centralized that the government at the top decides everything for the entire nation.
09:01
So yeah, so we also provided the county level so that if the road agency in a specific county can just see what he needs say for Nairobi County alone and also the specific constituency. So yeah, so that was it. And then now also one last functionality
09:21
on this interface was also to show if say they chose to maintain a road, say this road section here E, it would show an alternative route which is the blue highlighted route on the end. So if the road is closed for maintenance,
09:41
then it means transport for people and goods don't be distracted, meaning they can still use alternative routes to move from one point to another. Yeah, and then now the last dashboard was just an interface to show tweets as they come in.
10:01
So basically it would show the total number of reported accidents and the location of the tweets as they were being sent. So we were tracking for accidents, for road closures, maybe for issues that are not pertaining maintenance by this agency.
10:21
So maybe if there's a road closure because of something else, yeah. And potholes that have been reported also through Twitter. So those tweets are just tweets that we used for the pilot. And then now what we are currently working on is creating an analysis dashboard
10:42
where they can basically upload, say, a file of the roads, if it's a GeoJSON or a SHEP file or anything, and then they can indicate the maximum amount of roughness index, because that's what they use to measure
11:01
the condition of a road, and then it would show them a list of the roads that are in the worst condition. And then since Twitter also disabled the geolocation component, for now you can't share exact GPS coordinates for a tweet unless you geotag. And as we had seen,
11:21
most of the people who treated incidences were not attaching images. So also looking into natural language processing for Twitter feed to see if we can also be able to actually convert text to location. And then you're also exploring state level imagery through Mapillary.
11:40
Yeah, and that's it. Thanks, Laura, does anybody have a question?
12:08
Hello, thanks. If you implemented a linear reference on your old network, was it successful with open source tools?
12:24
Yes, I think as you had seen, only one section that is missing is for the first interface over there, it was on the back end there's an actual analysis that was being carried out for you to get this reading from A to E
12:42
and all the tools that are used are open source. Yeah, so all the processing was being done on the server side. We have another question.
13:03
You talked about the road agencies in each of the districts or counties as you call them in Kenya. Are the agencies committing data to open street map? Are they on board or is it still disparate
13:22
in terms of so that you can actually get a national data set for Kenya that is actually up to date, particularly for rural roads? So no, yeah, as it would have been expected, so basically for open street map it's the local community
13:43
that actually does a lot of work and then reach out to other organizations that have data on transport and then we sort of get to integrate the best data and what you get from these organizations. But actually for the first base of the road transport that was ever put on open street map,
14:02
they gave it to us, but then for the updates, no, yeah. Any more questions? So my question is maybe because now everyone has mobile, this is a question, maybe some suggestion because nowadays almost everyone has a smartphone
14:23
and you develop already applications, so maybe, and these smartphones, most of them they have accelerometer, so maybe you can integrate these features that when person goes in the car, that accelerometer automatically detects some shakes higher, for example, in some magnitude and you already automatically will know
14:41
that it's a bit hole in the road, yeah. Yeah, that is actually a solution that we had explored to integrate, but we were not looking to have the users to have to install say something else, but actually use what they're currently using,
15:01
something like Twitter on a daily basis to tweet normally because for that I think we'll have to make them install something or that would send out that information to our end, so we are trying not to change the user behavior and just use what they're currently using. We'll require some more permission from the user
15:23
to accept, for example, that application needs success to some more data from mobile phone, but I don't know if from the point of view of Kenyan legislation, it will somehow be, yeah, you need to think how to, basically possible to anonymize this data
15:42
because I understand it's some issue from point of view of personal data if someone tracks where you go, but somehow I think it's possible to think about some anonymizing algorithms for this. Okay, thanks. Any more questions? Okay, let's give an applause to Laura.
Recommendations
Series of 24 media