1 of 45

RAMANI Galileo Location-API (Android)

�An introductory

Seamlessly work with (uncertainity in) Global Positioning Data (GNSS) for robust Location-aware Smartphone Apps

RAMANI, an Ujuizi HOLDING B.V. company

Rondweg 40 M: +31 6 29072718

7468MC Enter S: valentijn_venus

The Netherlands E: valentijn.venus[at]ujuizi.com

2 of 45

What we do…

Launched in

2012

Initiated by

ESA

> 60

Apps developed

> 500

Participants

39.843

Hours of

Coding

AZO and RAMANI have organised annual Space App Camps for ESA since 2012 to foster the uptake of Copernicus data in mobile apps, which has also seen emerge new start-ups and commercial services.
Developers from all over Europe were invited to use the tools and data provided by the RAMANI and give feedback. A technical help desk furthermore addressed any problems or difficulties developers had while using the Mobile App SDK after these event to help grow Europe’s startup community be succesful.

3 of 45

Bridge the digital divide...

Aside a few exceptions, the majority of solutions to access global positioning satellite data focus on serving specialized hardware solution rather than helping normal App developers to improve location accuracy.
While the list of commercial support for EGNOS and GALILEO is growing, many promises of such support is lagging behind with reality (particular in the fragmented Android ecosystem). For example, the Huawei Mate 9 has issues to correctly receive GALILEO signals unless configured correctly:

snippet GPS configuration parameters in /vendor/etc/gps4774config.xml

Our Location-API bridges the “digital divide” between producer and consumers of location-aware Apps allowing them to tap into Europe's GNSS programmes.

4 of 45

Location-API�Key features:

Inform user when GALILEO and EGNOS Satellite-based augmentation systems (SBAS) are being used in improving location-accuracy (i.e. ionosphere correction)
The Location-API shows with/without Europe’s GNSS satellite systems to emphasize the added-value (time-to-fix, location-accuracy, etc.)

5 of 45

Location-API

A custom LocationManager for Android developers, which allows them to use advanced positioning services for location-aware Apps, such as:

Using inertia sensors and advanced sensor fusion algorithms, support GPS-less positioning to support complex user scenarios (e.g. travelling through a long tunnel, inside buildings, etc.)
Supports offline, vector-based background maps, always up-to-date sourced from open-source data collection efforts
Turn-by-turn instructions for navigation
GeoLocation services for address to coordinate translation
Crowdsource between road users (e.g. to report bad road pavement conditions). Some examples of crowd-sourcing in the next slides...

6 of 45

Location-API�Key features:

All exposed through an easy to use API-library:

GPS signal (SNR) bar chart, showing the signal strength for each satellite, as well as the accuracy and status of the GNSS network.
Satellite positions in sky-view
Other location-based information, such as current speed, heading and altitude, as well as the current time read from the GPS and the local time in your current timezone, as well as the sunrise and sunset times at your location.
Filter between GLONASS, GALILEO, SBAS, BEIDOU and QZSS satellites

7 of 45

Routing-API�Key features:

Turn-by-turn Navigation
Search Address by Name and by Point
Offline and Online Routing

8 of 45

Routing-API�Key features:

A custom Navigation implementation for Android with main features are :

Turn-by-turn navigation showing list of turns along the route (roads and signs to turn)
Search Addresses by name or by Point
Offline Routing, generate turn-by-turn routing without internet connection
Online Routing, generate route using our RAMANI service and custom weighing (e.g. incl. traffic info)
Snap-to-road (see next slide)

9 of 45

Routing-API�Key feature “snap-to-road”:

Enable snap-to-road using OpenStreet Map data (when poor geo-location provisioning inherent to certain smartphone user scenarios can be addressed by ‘snapping’ the received GPS-fix to the nearest infrastructure, e.g. road or pedestrian track)
Needed for crowdsourcing, because trade barriers can only shared and monitored over time if registered correctly on the road (and not on the roadside)

As illustrated above, we address poor GPS-signals on the device at run-time by snapping the coordinate reported by the smartphone’s GPS to the roads infrastructure provided by OSM.

Note that the accuracy of the reported location, according to Google’s software framework, is reported as the radius of 68% confidence.

If this location does not have an accuracy, then 0.0 is returned. All locations generated by the Android’s LocationManager include an accuracy.

So basically, if it returns 7.5, that means there is a 68% probability that you are within 7.5 meter of the GPS location reported.

In other words, if you draw a circle centered at the reported location's latitude and longitude, and with a radius equal to the accuracy, then there is a 68% probability that the true location is inside the circle. In statistical terms, it is assumed that location errors are random with a normal distribution, so the 68% confidence circle represents one standard deviation. Note that in practice, location errors do not always follow such a simple distribution.

Note that this accuracy estimation is only concerned with horizontal accuracy, and does not indicate the accuracy of bearing, velocity or altitude (even if those are provided by the Android Location provider).

10 of 45

How can better positioning help value-chain players?

Monitoring Road Infrastructure�(reduce maintenance time/cost)
Creating Platforms for Reporting Complaints / Harassment�(expose corruption)
Help Connect Buyers, Sellers and Service Providers�(generate sales cost-effectively)
Make People and Businesses Easier to Locate�(improve delivery time, reduce cost)
Transport match-making services (incl. blockchain ‘smart-contracting’)

(reduce idle transports, reduce CO2, build trust so value-chain players transition more)

Location-API

11 of 45

Demo app

Download the Application on any supported device using the following Google Play Store link (click on the Google Play picture below):

Location-API

12 of 45

SDK Version

Import library on Project/root build.gradle

Location-API

Import library on app build.gradle

13 of 45

Sample Code Reading Satellites/GNNS Status Data

Location-API

14 of 45

Sample Code Setup Routing Data

Location-API

Sample Code Setup SnapToRoad

15 of 45

Sample Code Setup Mock Location

Location-API

16 of 45

API for Earth Observation data

Mobile-ready technology to develop meaningful apps

17 of 45

Maps-API�Key features:

With our latest Maps-API, you are able to use a 3D Map view using OpenGL vector-tile rendering.

18 of 45

Ramani’s Maps-API�Key features:

A custom Maps implementation for iOS and Android developers, which allows them to use advanced map data services for location-aware Apps. Main features are:

Supports *offline, vector-based background maps (always up-to-date sourced from open-source data mapping efforts)
EO data is streamed just like video (block download is only optional)
data analytics platform helps to create added-value products using semantic workflows and visual IDE
author-reviewer-publisher roles that supports collaboration with other users and experts in ensuring data quality
supports multiple platforms is provided by native smartphone ecosystems (iOS, Android) and Web (e.g. JavaScript and Node.js).

*Android only

19 of 45

Combine Location-API with Earth Observation data�Mobile-ready technology

EO-data and other added-value Copernicus data products anywhere with our mobile-ready Maps-API
Integrate EO-data intelligence directly into your mobile application effortlessly and with no compromises
Address customers needs directly with highly detailed, regional EO-data on any mobile platform

20 of 45

Example Apps making use of our APIs�Mobile-ready technology

CHEETAH
Walkability
All apps developed at the ESA Copernicus hackathons and App Camps (+50)

21 of 45

PROMOTING EUROPEAN-AFRICAN (INTER)TRADE FLOWS:

THE ROLE OF CROWDSOURCING COMBINED WITH SATELLITE TECHNOLOGY

22 of 45

Doing business in Africa is very costly...

Long delays at ports and borders
Non-recognition of documents or certificates beyond national borders
Harassement along transit corridors, mainly from uniformed services
Cumbersome Procedure
Unfriendly Business Environment
Corruption

23 of 45

Volume of African Intra-Regional Trade

Intra-African Trade as % of total trade (2015)
REC	% of total exports	% of total imports	Intra-REC Trade (% of Total Trade)	Position
SADC	20.9	20.6	20.75	1
EAC	18.8	7.2	13	2
ECOWAS	10.8	9.3	10	3
IGAD	15.5	3.1	9.3	4
COMESA	12.2	5.9	9	5
CENSAD	6	5.9	5.95	6
AMU	4.4	3.2	3.8	7
ECCAS	1.8	5.2	3.5	8
Africa	17.7	13.6	15.65

Source:

AfDB, 2017

24 of 45

How can better technology help value-chain players?

Monitoring Road Infrastructure�(reduce maintenance time/cost)
Creating Platforms for Reporting Complaints / Harassment�(expose corruption)
Help Connect Buyers, Sellers and Service Providers�(generate sales cost-effectively)
Make People and Businesses Easier to Locate�(improve delivery time, reduce cost)
Transport match-making services (incl. blockchain ‘smart-contracting’)

(reduce idle transports, reduce CO2, build trust so value-chain players transition more)

RAMANI’s APIs for Crowdsourcing and EO data

25 of 45

26 of 45

Crowdsourcing value-chain intelligence

play introduction video

Sources: Venus et al., 2014; USAID, 2008

Frequent delays

Bribes

Rough roads

In poor countries, predominantly in Africa, an estimated 40% of food is lost at post-harvest and processing stages. Poor infrastructure, processing & packaging and un-refrigerated transport & storage are largely to blame. Postharvest losses are further aggravated by (illegal) toll stops and police controls as they prolong a journey and put profit margins under even more pressure. Figure 1 shows road obstacles and their impact on transport conditions for three West African countries (modified after USAID, 2012).

Each day scores of lorries laden with pineapples, mangos, tomatoes, onions and other fresh vegetables start the 1,000 km trip from landlocked Burkina Faso to Accra, Ghana’s capital on the Atlantic coast. This journey—along smooth & tarred roads in the north and bumpy & unpaved roads in the south—should take between seven and 12 hours.

By the time it gets to market, often two days later, most of its produce is perished. Produce decomposes fastest when they travel on bumpy roads over long periods in environments that are hot, dry, sun-drenched or stuffy. Under these conditions, the spoilage rate varies between 50% and 100%. But when they move over smooth roads through tall forests that provide shade and coolness, as well as humid air with plenty of circulation, only 25% spoil. And if the trip takes one day instead of two, the losses drop to 15%.

Environmental conditions along the trade corridor, particularly those known for their stress reinforcing impact on the cargo (Venus et al., 2014), become even more important given the un-refrigerated modes of transport used in the region.

Climate change has already started to affect African countries’ efforts to provide better transport services to their citizens. African transport systems are vulnerable to multiple types of climate impact: sea level rise and storm surge, higher frequency and intensity of extreme wind and storm events, increased precipitation intensity, extreme heat and fire hazard, overall warming, and change in average precipitation patterns. The increased frequency and intensity of extreme climate event challenges the year-round availability of critical transport services: roads are damaged more often or are more costly to maintain; expensive infrastructure assets such as ports, railways or airports can be damaged by storms and storm surges, resulting in a short life cycle and capacity than they were originally designed for. Critical infrastructure such as bridges continue to be built based on data and disaster risk patterns from decades ago, ignoring the current trend of increased climate risk. For Sub-Saharan Africa alone, it is estimated that climate change will threaten to increase road maintenance costs by 270% if no action is taken.

27 of 45

CHEETAH��Crowdsourcing barriers to trade and post-harvest losses in logistic channels

Indicators

27

Technical impacts
Food spoilage reduction		Up to 40%
CO2 Reduction		Up to 50%
Less OPEX		Up to 35%

Socio-economic impacts
Target group	+1 Mln	Users per country
Downstream opportunity	10	SMEs per country
Investment for SME	0-6K	Free of charge (open-data policy)
Employment	5K	# direct jobs per country

Countries

Burkina Faso, Ghana, Mali, Senegal, Togo, Cote d’Ivore,

Kenya, Tanzania, Zimbabwe, South Africa, Mozambique, etc.

Involved partners

Dutch: Utwente, Ujuizi Laboratories NL

Local: Borderless Alliance, Ujuizi Laboratories (Ghana, Kenya, South Africa), South African Trade Hub

28 of 45

Connect

29 of 45

Crowdsourcing road conditions

Road pavement quality has big impacts on truck maintenance and operational costs (example video)

Road pavement maintenance can boost local livelihoods though Labor-intensive Public Works (subcontracting)

Road Maintenance

Subcontracting

Sources: Rijn, 2005; Worldbank group, 2014

Crowdsourced

30 of 45

Smartphone app using AI and (human) sensors

Map-view

Sensor-view

31 of 45

Inventorying Road pavement quality, Accra region

legend

32 of 45

Track&Tracing Post-harvest Losses

33 of 45

Getting started

Try our on-screen demos in the App at the Google Play store

34 of 45

Crowdsourcing

To guide (semi)illiterate smartphone users to crowdsource, aka share, all sorts of location-specific information, i.e. the price of a commodity, the market name, quality, packaging type, etc., using swipe-based pictographic interactions usable for all age-groups and literacy levels..

35 of 45

Crowdsourcing commodity prices in emerging BoP markets

sELECT COMMODITY

sELECT MARKET

SET COMMODITY PRICE

Confirm

REPORT

Low cost agricultural market information systems by linking data to traders’ bookkeeping

Murekatete, a female street vendor in Rwanda, lives in Masaka but buys yellow bananas, avocadoes and other produces from Kabuga to sell in Remera, a borough of Kigali. Margins are thin. On a good day, she can make Rfw 2,000 (2 $), - if she walks. Without volume, she cannot afford a taxi. If you have to walk much of the day, it is important that you walk in the right direction to the market where you can fetch a reasonable price and that you bring the produce which is in demand.

For women like Murekatete, access to a market information system would be crucial to decide every day what to buy and were to sell. A wrong decision might imply that the two daughters of the single mother would remain hungry. Millions of women in Africa would benefit immensely from good market information systems. For women engaged in cross-border trade, who often have to cover even larger distances, such systems would be even more important.

In a recent contribution in Bridges Africa on Empowering Women by Supporting Small-Scale Cross-Border Trade, Emma Marie Bigingo, the Executive Secretary of Pro-Femmes/Twese Hamwe, dedicated to women empowerment, has argued that future interventions to improve the situation of women in cross-border trade should make use of advanced technology, especially mobile phones. Almost all their beneficiaries use phones as their main channel of receiving and communicating market information.

At the start of this century, there was great enthusiasm in many African countries for market information systems, which did provide farmers and traders with up-to-date commodity information. They could access daily fruit and vegetable prices from different markets through a text message, so they could head to markets safe in the knowledge of what to expect for their produce.

However, many of these initiatives were short-lived, like the Eastern Corridor Agro-market Information Centre (ECAMIC) in Ghana, or the Kenya Agricultural Commodity Exchange (KACE). The Forum for Agricultural Research in Africa (FARA) has published an Inventory of Innovative Farmer Advisory Services using ICTs of these early initiatives. Reasons for their low survival rate include that it was very labour intensive to gather the information on prices of several markets and pass this information on in text messages. Many farmers did not have a mobile phone yet, and the coverage of networks was not as wide as it is today. In the meantime, much has changed.

According to GSMA study The Mobile Economy. Sub-Saharan Africa 2018, overall subscriber penetration in 2017 reached 44 percent, up from 25 percent at the start of the decade. Even remote areas are now often covered by mobile networks, and prices for calls and text messages have dropped in most countries. At the same time, opportunities to store data in the cloud and share them with other users have developed at high speed. That offers very different opportunities for systems that provide access to real time information on the prices of different products in various locations.

Using these new technologies, there have been many new initiatives to collect price information and disseminate it to users. The biggest challenge for these initiatives are not of a technological nature. The biggest challenge is to collect reliable up-to-date information on a daily basis at low cost.

Research teams from the Joint Research Centre of the EU in 2017 published two studies on alternative data collection methods, one on Using web and mobile phone technologies to collect food market prices in Africa. Approaching real-time data and use of crowdsourcing, 2013-2016 and the other one onInnovativeFood Price Collection in Developing Countries. Focus on Crowdsourcing in Africa.They investigated the potential of new technologies and innovative data collection methods through citizens’ contributions (professional or amateur). The challenges they encountered included ensuring crowd participation, choosing the right technology, and obtaining data of sufficient quality. Financial issues like cost-efficiency and long-term sustainability also formed a big problem for most initiatives.

One of the study provides a detailed description of presently existing market information systems that use modern ICT in African countries, including Burkina Faso, Ghana, Ivory Coast, Kenya, Madagascar, Malawi, Niger, Nigeria, and Togo. Some of these systems were introduced by state agencies, others by private companies or by non-governmental organisations. Their costs usually cannot be covered by the service and are financed by a donor agency. A broad range of institutions has financed such initiatives, among them the World Bank, the EU, the World Food Programme, the African Development Bank, CTA, USAID, ECOWAS, Mercy Corps, Catholic Relief Services, Oxfam, and Save the Children.

Data collectors of most systems are paid or receive a non-monetary incentive. This reaches up to 15-30 US-$ per market surveyed and up to 120 US-$ per data collector. The high costs of data collection are the main reason why all reviewed crowdsourcing information systems struggle with financial challenges.

To make market information systems viable and independent of continuous support from donor agencies, the running costs have to be brought down. One way to do this would be to go a step further than the initiatives reviewed in the two studies and work without special data collectors. Instead, the information could be linked to real life transactions, which would directly be used as an input for the information system.

Such a system could operate in the following way:

A sufficient number of traders is equipped with smartphones or tablets and trained to use a simple bookkeeping programme that keeps track of all income and expenditures. Traders have to do that anyway to remain successful in their business. The data remain private. However, the product purchased or sold and the unit price for that product would be forwarded automatically to a central data base. The device can include information on the time and the location of the purchase, if the transaction is keyed in immediately after it took place. The only thing that the traders would have to do extra is to note also the unit price and not only the total cost of the purchase.

At the central data base, the average price for each product on that day would be calculated on the basis of all prior entries. The system could thus even keep track of price changes during the day. It would give other traders real-time information about prices at different market places, without going there themselves, if any other trader, participating in the data entry, had sold or purchased a product of the same category on the same day. Technically, it would not be difficult to devise such a system.

But there are, of course, a number of trade-offs to deal with: On the one hand, the system should be very easy to handle for both, the traders who contribute by their own bookkeeping to the data in the system, and the traders who only use the system.

The contributors, for example, would only see the products that they normally deal with on their screen, e.g. bananas, onions, and tomatoes. They would just have to click on a little picture and add the unit price. The system could automatically detect improbable outliers and leave them out.

The users would not need a smartphone. They could access the data base with a simple phone and would only receive the prices for products they are interested in, on markets within their reach. They can easily adapt the list of relevant products and markets whenever they wish. The data base could also be connected to an interactive voice response system, as in Malawi, which allows users to dial in and hear the prices.

A big challenge is the selection of the different product categories for which price information is collected. Even for the same products (e.g. tomatoes), prices can differ considerably in the same market place, depending on the variety, quality, size, freshness, time of the day, and location on the market. So, the information on the average price might not tell enough, if for the traders, who depend on very small margins, already minor differences are important. A system linked to the bookkeeping of traders would not allow to add qualitative information to the prices.

What could convince traders to link their bookkeeping to such a market information system? In the first place, it would help them to get access to a simple bookkeeping device that would help them to improve their business. Their data would remain private, - only the product categories traded, and their unit price would be passed on. Since the data are part of their own bookkeeping systems, they have an interest to keep them accurate. They would get a smartphone or tablet and a training in the good use of the device and the bookkeeping programme. In addition, they would get airtime to send the data immediately to the central data base, which could happen automatically. In Burkina Faso, for example, Sonagess has created a virtual private network accessible only by the collectors’ smartphones. It is a paid service, preventing the need to pay individual internet fees for each agent. Unfortunately, the software is proprietary and not available to the public, but a similar scheme could also be introduced elsewhere. Contributors might also get additional airtime for private use. In Kenya, for example, contributors are compensated for each entry with airtime for 10 text messages. In addition to airtime, the contributors could also get regular feedback on their bookkeeping, with suggestions on how to improve their business further. They could also be invited to an annual conference to refresh their skills, exchange experience with others, and get informed about new features.

The system will only work in regions where the internet is accessible most of the time. Many of the existing schemes struggle with considerable downtime and have to fall back on SMS-messages in periods without internet access. The suggested system would be most suitable to countries with a relatively stable internet connection, like Rwanda, where 4 G network coverage now has reached 95 percent of the country’s territory.

Such a market information system, once installed, could be run at relatively low cost. It would offer precious up-to-date information for traders which would help them to waste less time and to increase their income. It would also be a valuable source for public policy making. It would deliver insights into the development of food prices in different regions, shifts between different commodities traded, and could contribute to an early warning analysis when food security is endangered.

36 of 45

Example Market Trends Ghana

37 of 45

API for Earth Observation data

Mobile-ready technology to develop meaningful apps

38 of 45

What others have to say…

Niels Rasmussen

Senior Associate, Trade Facilitation

Saana Consulting

www.borderlesswa.com

39 of 45

What others have to say (cont)…

Luke A. Ney | Program Analyst, Millennium Challenge Coordination Team | E3/TRR | USAID

lney@usaid.gov

Ziad M. Hamoui

President,

Borderless Alliance

www.borderlesswa.com

40 of 45

Understanding Sentinel-2’s Multi-spectral Imager (MSI) for road pavement (quality) detection

41 of 45

How can better technology help value-chain players?

Creating super-resolution data from freely accessible EO observations

Worldwide coverage in 10 m resolution and full colour, very recent, updated to every 5 days. However, resolution of freely accessible EO data (e.g. Sentinel-½) is in its raw format only sufficient to trace major roads and landuse features. But, thanks to the simplicity of road pavement as a physical structure we apply spectral lineair unmixing and a Deep Neural Network to map at sub-pixel level the type, age, and conditions of road pavement

Creating value-chain monitoring products with EO data

Based on the above, we have ongoing R&D efforts to map delays at border crossings based on EO (e.g. length of queues based on Sentinel-½ timeseries validated with phone data collected through CHEETAH)

RAMANI’s APIs for Crowdsourcing and EO data

42 of 45

RAMANI is

3 locations worldwide
6 FTE with a combined +35 year experience in Web 3.0 technology and Geospatial data

RAMANI’s mission is to contribute to the advancement of public use of geo-spatial data, particularly by producing easy-to-use toolboxes (in the form of so-called Application Programmable Interfaces – APIs) to facilitate the up-take of geodata through advanced technologies stacks (GeoICT, AI, blockchain, etc.). This lowers the entry barrier for SME ICT companies, with average skill-levels, to also integrate and apply them in regions still subject to intermitted, poor (mobile) Internet access (BoP). Using the cloud platform RAMANI Analytics (https://analytics.ramani.ujuizi.com/) all sorts of GeoData can be easily retrieved, analyzed, and published to mobile and web; it’s development cycles till date, based on investments form public and private partners, represents a value of +1 million USD. Ramani has +4 years of experience as key-technology provider of some of Europe’s largest innovation competition events, such as the Copernicus masters and various sponsored app developer camps, fully funded under contract with ESA.

43 of 45

Launched in

2014

> 60

Apps developed

> 500

Participants

39.843

Hours of

Coding

Jul

Aug

Sept

Oct

Nov

Dec

…

July

Aug

Sept

Oct

Nov

START EVENT

STARTUP

PHASE

AWARDS CEREMONY

Dec

Recruitment and training of the candidates

Awarding of all winner(s)

Shortlist for the AppCamp

Provision of acceleration and incubation services & syndication

Timeline of an typical innovation event (and RAMANI’s role as technology provider)

44 of 45

Value-chain Tracking Services (CHEETAH)

Road quality info from Edge-AI and EO helps

to alarm road users & road maintenance planning to be better targeted.

Bribes and delays eat away revenue of small traders.

Crowdsourcing of e.g. Market Prices offers novel means to collect data but also to exact accountability from Non-Tariff Barriers.

Post-harvest losses are quantified as a function of trip duration, vibrations, and satellite-retrieved microclimate (Sentinel-4).

45 of 45

RAMANI – spin-off Utwente

Ujuizi HOLDING company

Rondweg 40

7468MC Enter

The Netherlands

Valentijn Venus (CEO)

+31 (0)6 290 72 718

valentijn.venus@ujuizi.com

Explore

Inspire

Connect

Invest

Make Technologies Work Innovators

Open Work Space
Digital Education
Heros

Innovation Ecosystems
Startups, Innovators
Experts

Public-Private Partnership (PPP)
New Space Startups

Team up!