16 aprile, 2012 | di Lorenzo Perone
Questo articolo è stato pubblicato originariamente sul Blog di Working Capital.
Ushahidi, che in Swahili significa testimonianza, è oggi una una società non-profit che sviluppa software libero. Il suo core business è lo sviluppo di software per la raccolta, la visualizzazione e la rappresentazione di informazioni su cartografia interattiva.
E’ un valido esempio di come una startup possa nascere da un’idea di successo, in una realtà come quella del Kenya, che non associamo a incubatore di tech company.
Tutto nasce durante le violenze che si sono verificate in Kenya, nel 2008 a seguito delle elezioni politiche. Un gruppo di persone tra cui l’attivista kenyota Ory Okolloh, si trovarono a documentare sui loro blog, anche attraverso i commenti, le violenze che stavano accadendo nel paese. Qualcuno lanciò l’idea di mappare i luoghi delle violenze attraverso Google Maps, e Ory rilanciò la proposta via twitter, chiedendo supporto alla comunità degli sviluppatori software del paese.
In tre giorni era online Ushahidi.com, il sito originale, ancora visibile in una sezione del sito attuale. Ushahidi ebbe una grandissima risposta, ben 45,000 utenti inviarono le loro testimonianze dai luoghi degli scontri, permettendo di seguire l’evoluzione della situazione nel paese in maniera indipendente da quanto riportato dagli organi ufficiali. Un esperimento riuscito di citizen journalism e crowdsourcing.
Da allora Ushahidi non rappresenta più solo una piattaforma web ma una realtà aziendale complessa, al cui al nucleo originale dei fondatori Erik Hersman, Juliana Rotich, Ory Okolloh eDavid Kobia si è affiancato un nutrito gruppo di sviluppatori e di specialisti di tecnologia, attivisti politici, giornalisti ed esperti di comunicazione. La cosa che colpisce, guardando le pagine del team è l’estrema distanza geografica che separa queste persone che vivono, in qualche caso, in continenti diversi. Il cuore pulsante del progetto rimane in Kenya anche se, formalmente, la società ha sede ad Orlando in Florida.
La piattaforma di Ushahidi dal 2008 ad oggi è stata testimone di numerose emergenze tra cui ilterremoto di Hahiti, la crisi libica, il terremoto in Giappone e l’attualissima crisi siriana, riuscendo a supportare i cittadini e le organizzazioni che operavano sul campo attraverso la gestione e pubblicazione su mappa delle richieste di soccorso e delle risorse disponibili.
La facilità nella diffusione è stata favorita da Crowdmap, uno dei prodotti creati dalla startup a partire dall’idea originale e che permette di utilizzare le risorse web di Ushahidi per mettere on line, gratuitamente, una versione personalizzata della piattaforma in pochi minuti.
Questo ha permesso, anche in Italia, la realizzazione di diversi deployment, tra cui Rifiutiamoci.
Uno dei punti di forza di Ushahidi è senza dubbio la versatilità. Consente di operare in condizioni logistiche di comunicazione profondamente diverse, adattandosi a scenari di utilizzo profondamente differenti.
La funzionalità di base è quella di rendere disponibili, anche in tempo reale, informazioni geolocalizzate relative a diverse categorie di eventi su una mappa, permettendo di analizzare il flusso temporale degli avvenimenti attraverso una timeline che può essere attivata per rappresentarli dinamicamente.
E’ possibile ricevere via email o SMS un messaggio se un evento di determinato tipo si presenta ad una determinata distanza da una posizione prefissata, allertando ad esempio un giornalista o un’unità di soccorso.
Le segnalazioni possono essere inviate via SMS o nel caso di uno smartphone con GPS integrato, utilizzando il formato GeoSMS che invia automaticamente indicazioni rispetto alla propria posizione. Sono state create specifiche apps per iPhone ed Android ed è prevista un’interfaccia dibackoffice per inserire le segnalazioni ricevute telefonicamente oltre che per validare le diverse segnalazioni, provenienti anche da email e da web.
Un’altra delle innovazioni legate allo sviluppo della start-up è stata la realizzazione di una innovativa piattaforma: SwiftRiver, che è attualmente in versione beta, che analizza e verifica inreal-time grandi flussi di dati provenienti da diversi canali quali: email, twitter, SMS e feed RSS. Il sistema provvede ad un’analisi semantica del flusso permettendo la l’auto-categorizzazione e classificazione dei dati analizzati sulla base di parole chiave.
SwiftRiver è utilizzato anche da Wikipedia per analizzare come gli editor della famosa enciclopedia tracciano, valutano e verificano l’attendibilità delle fonti relative alle voci pubblicate.
Ushahidi ha maturato una così grande esperienza sul campo da aver creato anche dei veri e proprihow-to per l’organizzazione di un sistema di monitoraggio elettorale, per il fund raisinge per la gestione della sicurezza in realtà “democraticamente” difficili.
Tra i progetti italiani che lo utilizzano possiamo citare:
10 aprile, 2012 | di Antonio D'Argenio
A short story
Some years ago I was working for my company together with my colleagues Andrea and Fabio. That was one of the rare times we had some money to use to integrate our equipment. We work in GIS, mostly web-mapping and numeric cartography, and at that time we were lucky people with new graphic workstations, a small sized but new brand server farm and all we needed to do our job at the best. We had just completed a project related to some environmental issues and one of the problems we collided with was the lack of geodata at the scale we were looking for. At that time, public Italian geographic data, especially those referred to the region where we live and operate, were not so easily available nor frequently updated. This is a well known problem for people working with GI so Fabio told us “Why don’t we buy a small airship to collect our own information?”
Starting from this moment we spent about one month looking for the right aircraft.
At the beginning we had rough and confused ideas of what to do. Obvious constraints were: affordability, reliability and ease of use. But we were also looking for GPS control, flight planning and everything could be connected with mapping. With our surprise we easily found a number of carriers (not only airships but also planes, helicopters, paragliders, multi-copters and so on). Some solutions were of industrial type, other substantially homemade but we were looking also for a partner, someone who could help us to start and this item became part of our “requirements list”. We also found that our budget allowed to buy more than one aircraft so we decided to concentrate on two different kinds of vehicles: a paraglider to be used in photogrammetric surveys of large areas and a small quadricopter which we planned to employ on smaller areas. In those weeks we also found Daniele, a colleague and a friend that introduced us to aerial mapping with small, Remotely Piloted Aircrafts.
What sRPA are
RPA means Remotely Piloted Aircraft. An “s” at the beginning stands for small.
Remote control implies that the pilot is not sitting on-board but generally remains on the ground surface, controlling the flight using a radio device.There are a number of equivalent ways to call this kind of aircrafts. Probably the most common are UAV (Unmanned Aerial Vehicles) and drone (even if this last term originally indicated military targets
Since remote control is a base characteristic to define an RPA, there are a lot of crafts kinds falling in this category. Beside obvious differences (such as those between fixed and rotary wings) other important parameters to take in account are: size, weight, range, maximum altitude and endurance. These last parameters are those commonly used to classify unmanned aircrafts. The table below illustrates an old classification of the UAV Association that uses the above mentioned parameters to define classes and, even if not the most updated, can be used as guide.
||150 to 300
||10 to 30
||2 to 4
||30 to 70
||3 to 6
||70 to 200
||6 to 10
|Medium Range Endurance
||10 to 18
||50 to 9000
||0.5 to 1
||24 to 48
Small RPA are those falling within the micro and mini classes. If you look at the table, you easily understand why they are so attractive for professionals and SMEs working on mapping of small areas (up to tens of hectares). They are often so lightweight and small that you can carry them in the back of your car, but they can fly high and far enough to be employed to work. Furthermore they are enough safe. For instance: most of the small RPA are equipped with electric engines, which means that there is no fuel flying over your heads. Some RPA, such as multicopters, are VTOL (Vertical Take Off and Landing) aircrafts so they don’t need runways and can start to fly from very small surfaces.
On the other hand, it’s important to notice that these objects have specific limitations that you must never forget. You should never fly over people or lose the sight of your drone. Wind, obstacles (such as electric cables or poles), electromagnetic interferences are enemies. You should also remember that, even if most of these objects are smaller and lighter than models and ultralight crafts, when you use them to work they are considered in all respects aircrafts.
I’m not an engineer, so I cannot deepen into technical issues but there are few important tech notions that you must have in mind if you want to deal with small drones. The most relevant are those related to some subsystems that are barely common to every UAV: GPS, IMU, communication and guide systems.
GPS measures the drone position during the flight. Some small drones also use the GPS for safety issues. For example, if you lose the radio contact with him, the drone can use the GPS to maintain its position until communication is restored.
IMU stands for Inertial Measurement Unit. An IMU is an electronic system that, using accelerometers and gyroscopes, measures speed and attitude of the aircraft during the flight. All the information you receive about the yaw, pitch and roll angles of your UAV came from its IMU.
From left to right: Yaw, Pitch and Roll angles. These angles define the aircraft attitude and information about them is provided by IMU.
This information reaches you thanks to a communication system that continuously transmits flight data to the ground. Generally these data are received by a ground base station and displayed along with other telemetry data using a software application.
Since you need to pilot the aircraft, communication between you and your drone is always bidirectional. The drone talks with you through telemetry, you talk with him using a radio control. Some small drones are also able to fly autonomously since they are capable to automatically execute planned flights. This last characteristic is an essential requirement if you want use them for mapping purposes.
Application to civil purposes
RPA technology has been originally developed for military purposes. This is the reason why the word “drone” often evokes war scenarios. When I did my first web survey to choose which drone to buy, I was impressed by the number of military sites related to drones.
Nevertheless, civil use of UAVs, especially the small ones, is growing. This is essentially due to the reliability achieved by these systems along with their cheapness and ease of use (if compared with conventional manned aircrafts).
Besides the applications in which they are used as point of view, there are a lot of technical fields where the use of a sRPA is helpful. The table below lists some of them.
||Photographic surveys on buildings, bridges, dams, etc.
||Preventive inspection of energy infrastructures
||Monitoring of construction and mining abuses
||2D and 3D modeling
||Monitoring of utilities infrastructures
||Illegal dumping monitoring
||Prevention of critical events
||Energy performance of buildings
|Monitoring of construction sites and public works
In all these fields, your drone is not simply an eye in the sky but can be used to collect measures. If you use a camera as payload, you can collect the right number of images to realize a photogrammetric survey. If you change the camera with a multispectral device you can collect Remote Sensing data. You could also use active sensors, such as LIDAR, but due to the low load capacity of small UAVs (typically ranging between 0.2 and 5 kg) and high power requirements, is not so easy to find the right payload.
Spatial information is a pervasive content and geomatics entered in different ways in most of the aspects of our life. With reference to the technical use of small drones, geomatics is both a way and a scope.
As I said before, flight plans are important if you want to retrieve data for mapping (e.g. photogrammetric data) and spatial tools can be efficiently integrated in flight plan design process. This happens because flight plans are based on carrier (and target) position, so all the details of the flight plan (route, acquisition points, etc.) can be correctly set or calculated starting from a GIS environment. Better: the whole planning process could be confined within a dedicated geospatial platform. Even better: geospatial applications represent the right environment to display and analyze most of the data you have collected.
Since when you plan a flight is because you have some goal, this goal shapes planning activities. If I want to measure the concentrations of some chemical compounds, probably I just need to define flight route and altitude. If I want to realize a photogrammetric survey, my needs will extend to camera parameters or images overlap and sidelap.
When you define Overlap and Sidelap of a flight plan you are setting the front and side superposition of frames.
These examples are not randomly chosen, because both can lead to mapping:
- spatialized chemical measurements easily lead to thematic mapping,
- photogrammetric survey can be used to generate orthophotos.
It is important to know that not all the drones allow automatic flights and not all the drones allow the same degree of automation. For example, our quadricopter is a very sophisticated tool that permit to define all the flight details (takeoff and landing points, route, trigger points or trigger frequency, camera zoom, camera pitch and roll angles and so on). On the other hand, our paraglider just allows to define 12 waypoints and we are forced to manually trigger the camera (is up to you to decide which is my favourite …).
Geomatics & Photogrammetry
Photogrammetry is probably the most diffuse technique used in combination with small drones.
In my experience, the photogrammetric approach is a swiss knife for a number of application fields. You can use variations of the same basic techniques to realize metric images and 3D models of the ground surface, models of buildings, bridges, quarry fronts.
If you want to use aerial photos for mapping purposes, one of the key issues is DSM generation. Digital Surface Models are three-dimensional representation analog to DTM. Differences between them are related to the inclusion (DSM) or exclusion (DTM) in the model of objects, such as buildings, located on the ground surface. Since to produce orthophotos you need to project your images on a model of the ground surface, if you want to employ your aerial pictures to generate orthophotos, you need to pass through a DSM (the good news are that, if you have collected your pictures following the right scheme, you have all the information needed to DSM generation). Once you have processed your images in order to obtain a DSM, there are other things you can do besides generating orthophotos. For instance, you can drape your images onto the model to obtain useful (and also beautiful) photorealistic 3D metric models.
Orthomosaic draped on a DSM. This image refers to an archaeological site in Western Sicily. From Borruso et al. (2011) Atti 15a Conf. Naz. ASITA, pp. 471-478
Models based on image processing can be integrated with those produced with other methods.
For instance, in some applications, coupling low altitude aerial photogrammetry with Laser Scanning, helps to complete high resolution and precision measurements of buildings with those parts (such as roofs) that are inaccessible from the ground.
In some workflows, you could need to integrate satellite data with data acquired with sUAV. This happens because of the higher resolution needed or since clouds mask some part of the satellite image or because you need to record a time series with frequencies not allowed by satellite return time.
Low Altitude Remote Sensing
Aerial images can be considered a special case of use of sUAV in the more general Low Altitude Remote Sensing (LARS) scenario so, even if photogrammetry is probably the most diffuse technique connected with sUAV usage, there is a lot of space for other employments.
Infrared imagery (both near and thermal) are good examples in this direction.
Near Infrared (NIR) is the region of the electromagnetic spectrum falling between 0.75-1.4 µm. A number of objects reflects (i.e. plants chlorophyll) or adsorb (i.e. water) the incoming solar radiation in this region. NIR data can be acquired using dedicated sensors or modifying commercial cameras
If you combine NIR data with images collected within the visible (VIS) range
TIR data can be used alone or in association to multispectral information to perform more complex analysis. For instance, you can use a thermal camera mounted on a UAV to measure energy performance of buildings or you can combine thermal and multispectral information to carry out surveys on illegal dumping.
Legal and tech issues
The examples we did since now represent the most common cases of use of sUAV for technical and scientific purposes.
The most natural question that we can do now is: if they are so reliable, quick, flexible and cheap, why sUAV are still relatively uncommon?
There is more than one answer to this question and the most relevant are related to some legal and technical issues.
The first critical issue is the poor definition of a general regulatory frame focused on the UAV use, that disadvantages also the diffusion of small drones. As we said before, these carriers are often lighter and safer than aeromodels or ultralight aircrafts but, even if these crafts can fly in conventional aerial space without limitations, sUAV cannot. This prohibition is related to their use. Roughly: if you do fly something with purposes other than your pleasure, you are doing aerial work and you must be subjected to rules. Most small companies interpreted the lack of rules as an implicit permit and this worked until small drones were unknown objects. This not correspond to the present situation, where an increasing number of potential players has tested the utility of light drones in the generation of services and sUAV increasingly appear on newspapers and TV. Due to this interest national and international authorities are trying to define a frame and a set of rules which recognize the special nature of these crafts.
Circular 328 – Unmanned Aircraft Systems (UAS)
European countries generally do not allow commercial use of small UAS with the exception of the United Kingdom, where Civil Aviation Authority (CAA) grants some permisions.
There are also a number of technical issues the first of which is the lack of dedicated sensors. Load capacity of small drones is a hard constraint that limits the employ of professional tools. Even if miniaturization of aeronautical devices is a constant process, most of the active sensors, such as LIDAR, are still too heavy to be mounted on a small drone. For this reason passive sensors, such as optical and thermal cameras, are the most diffused in daily practice. Other kind of tools, like electronic noses, even if already available are relatively rare.
Other topics belonging to this line are connected to future technical development that in a next future will improve the sUAV capabilities. Just to report a list of the most interesting:
- sense and avoid obstacles in autonomous flight
- ability to cooperate in formation flight
- increasing of endurance
- enhancing intelligence on board.
Communities and information
If reading this article has somehow intrigued, perhaps may interest you to know that there are a number of discussion groups centered on sUAS. If you want an access point, I suggest to write one of the RPAsynonyms in the group search of Linkedin.
There also associations focused on UAV. Among the most relevant, is important to cite UVSInternational, a non-profit association which includes unmanned vehicle systems (so, in this case, not only UAV) manufacturers, service companies and researchers. In these months it is also being founded the Associazione Italiana UAS which, as UVS International, is open to the all the players of the UAV chain that operate in Italy.
If you want to find more information on the topics covered in this article there are some specialized web sites that you can refer to. One of my favourite is suasnews.com which provides a daily updated information on UAS world.
Because of the strong relation between drones and geomatics, UAV news appear on various websites focused on spatial topics (try to search for UAV in GeospatialWorld, which also published an interesting review on UAVs).
Relation between UAV and Geomatics have been also the key topic of an international conference which was taken in Zurich in September 2011. The conference title was UAV-g 2011 – Unmanned Aerial Vehicle in Geomatics and here you can access to the proceedings.