GI Product Development Gerhard Navratil Workshop in connection

GI Product Development Gerhard Navratil Workshop in connection with Geomatics 2011, Teheran, May 14 2011

Contents • Motivating Examples • GI Product Design – Value of a GI Product – Data Requirements – Legal Aspects – Costs of a GI Product • Implementation? Gerhard Navratil Workshop GI Product Development

Motivating Examples • Course concept used for several years at – Vienna University of Technology – Surveyors – Technikum Wien (University of Applied Science) – Intelligent Transportation Systems • Results from student work Gerhard Navratil Workshop GI Product Development

Mobile GIS to Optimize Information Flow in Case of Emergency (1) • Fast reaction can save lives • Information in essential • Requires cooperation between organizations Use of new communication and information technology Philipp Nitsche Gerhard Navratil Workshop GI Product Development

Mobile GIS to Optimize Information Flow in Case of Emergency (2) Gerhard Navratil Workshop GI Product Development

Mobile GIS to Optimize Information Flow in Case of Emergency (3) Technical Concept TETRA maps CLIENT 5 § Terrestrial trunked radio XML, GML § TEDS: up toform 500 KBit/s job § For emergency- and rescue services § 2 -way authentification XML, CAP alert § Direct connection without telephone network possible XML, GML § Potential use of info different frequencies § Push-to-Talk Gerhard Navratil Workshop GI Product Development DB SERVER

Mobile GIS to Optimize Information Flow in Case of Emergency (4) Benefits: Gerhard Navratil Workshop GI Product Development

Emergency Calls for Elderly People • Increasing percentage of elderly people in Europe • Many living alone • How to help them when they need help? Susanne Pröstl Gerhard Navratil Workshop GI Product Development

Existing System 1 • Bracelet with sender • Connection to telephone line • Pushing a button creates connection to emergency response center • Handsfree equipment for ommnunication • If not possible: Person of trust informed Costs: Rent: 27. 50€/month Gerhard Navratil Workshop GI Product Development

Existing System 2 • Additional sensor around the neck • Automatic detection of fall • Automatic alarm Costs: 33. 90€/month Gerhard Navratil Workshop GI Product Development

Extensions • • • Built like a mobile phone Positioning by GSM-Positioning Relaying position to person of trust or ambulance Finding shortest route Communication via telephone But: Is it allowed to transfer the position to a third party? Gerhard Navratil Workshop GI Product Development

Personalized Traffic Information System (1) What is the current situation? • many traffic information services available – anachb. at – ÖAMTC – ASFINAG (road pilot) – ÖBB (Scotty), Wiener Linien (Qando) etc. a lot of information available but not personalized Peter Votzi Gerhard Navratil Workshop GI Product Development

Personalized Traffic Information System (2) • provides information depending on the user needs. • filters “unnecessary information” – users need only information affecting their trips – trip routes are often the same, e. g. , family visits tell the user only what he wants to know Gerhard Navratil Workshop GI Product Development

How does it work • user data needed: – trip routes from where to you go and when? – trip times how do you want to receive the information? – service channels • traffic events affecting the routes are transmitted to the user – route edges subtend affected edges – travel time exceeds a threshold GIS operations, functions needed Gerhard Navratil Workshop GI Product Development

What data are needed? • a common reference system (GIS) – road network, ways, points of interest, addresses • traffic information • travel time Gerhard Navratil Workshop GI Product Development

What functions are needed? • geocoding, routing, a map • service interface • service broker easy user interface provide the information receive traffic information, compute personalized results Gerhard Navratil Workshop GI Product Development

Where to get data? • Requirements – up to date, multimodal – can be used as reference system • ALERT-C/TMC, chain age (linear reference) – routing graph • commercial data providers? too expensive, old, not multi modal, license issues • solution: Open. Street. Map – up to date, open license (attribution, share alike) – Replication Gerhard Navratil Workshop GI Product Development

How do design? Architecture • • • Postgre. SQL + Post. GIS storage pgrouting, osm 2 pgrouting replication osmosis map rendering, OGC services UMN Map. Server client map Openlayers Apache Web Server, Asterisk, Sendmail web server, phone & sms gateway, email gateway all components are open source no license fees! Gerhard Navratil Workshop GI Product Development

Architecture reference system, user data, gis data Traffic Information Source datex 2 Database Postgre. SQL, Post. GIS push routing engine Service Broker Replication Website Map. Server Web Server SMS gateway Gerhard Navratil Workshop GI Product Development E-Mail gateway osm

Which standards to use ? • OGC – Web Map Service (WMS) – Web Feature Service (WFS) – Simple Features (database) • EASYWAY – Datex 2 or TPEG Gerhard Navratil Workshop GI Product Development

Tons of other examples • • • Tourist guides Support for Alzheimer patients Finding paths without crosswind Bus-stop optimization Fire protection system – are there still humans in a burning building? • … Gerhard Navratil Workshop GI Product Development

GI Product Design • What is the information content for the user? • What is the value for the user? • What data are necessary? • Access method? • How can the information be extracted? • What are the costs of this service? For the producer and for the user! Gerhard Navratil Workshop GI Product Development

Information Content Amount of data transferred through a channel can be measured (e. g. , in Bits – one binary decision) (Shannon and Weaver 1949) Useful to determine storage space etc. But is it useful for information content? NO: Measures the amount of data, not the effect of the message Gerhard Navratil Workshop GI Product Development

Example: Same Route Information? [1] Kirchberg am Wechsel to Gloggnitz • Follow the road to Otterthal • In Otterthal turn right towards Gloggnitz • Follow the road through Schlagl and Graben • Cross under the Semmering highway • Follow the road into the town of Gloggnitz Gerhard Navratil Workshop GI Product Development

Example: Same Route Information? [2] Your route from Kirchberg am Wechsel to Gloggnitz: The total distance is 13. 1 km. To drive this distance will probably take 00: 21 (hh: mm). • • 00: 00 On LH 134Markt 4, 1 km 00: 06 Turn right on LH 134Otterthal 6, 6 km 00: 16 Turn right on LH 134Graben 430 m 00: 16 Turn right on LH 134Graben 770 m 00: 18 Turn right on B 27Semmeringstrasse 650 m 00: 19 Turn right on Hoffeldstrasse 500 m 00: 20 Turn right on Sparkassenplatz 50 m 00: 21 Turn left on Sparkassenplatz 128 m Gerhard Navratil Workshop GI Product Development 4, 1 km 10, 6 km 11, 0 km 11, 8 km 12, 5 km 13, 0 km 13, 1 km

Pragmatic Information Content Shannon & Weaver: Transmission Assessment of information content must include the decision that is based on the data. Relative measure: better decision = more (better) information Frank: Pragmatic Information Content Gerhard Navratil Workshop GI Product Development

Necessary Discussion for a GIProduct • What is the information necessary for good decisions? • Level of redundancy? Gerhard Navratil Workshop GI Product Development

Value for the User (1) Products must have a benefit Benefit of GI-product may be (Krek & Frank 1999) • Reduced time to make a decision • Improved decision • Reduced risk Cost Problem: Cost of data increases with quality Gerhard Navratil Workshop GI Product Development Quality

Value for the User (2) Quantification of value? Value chain: The value chain concept according to Porter (1985) suggests that the activities in transforming raw materials and other inputs to final goods can be viewed as a collection of complementary and sequential tasks, each adding value to the final product. USERS Producer 1 data collector Producer 2 data integrator Producer N specialist Geoinformation product (Krek & Frank 1999) Gerhard Navratil Workshop GI Product Development

Costs in the Value Chain USERS Producer 1 data collector Producer 2 data integrator fixed cost high (sunk cost) Producer N specialist + Geoinformation product marginal cost = low, zero Gerhard Navratil Workshop GI Product Development total cost

The Value Chain Paradox USERS Producer 1 data collector low Producer 2 data integrator Producer N specialist VALUE Gerhard Navratil Workshop GI Product Development Geoinformation product high

Value for the User (3) How much is a client willing to pay? Maximum is determined by • the benefit • reduced by the costs of accessing the GI product Price = Maximum No gain for the user Price < Maximum Gerhard Navratil Workshop GI Product Development

What Data are Necessary Analysis of information leads to required data Example: Car navigation requires at least • Street network • Speed limit • Restrictions (turning, weight, height, etc. ) • Address information Missing data leads to services that do not work – how often do you specify your destination by coordinates? Gerhard Navratil Workshop GI Product Development

Why is Data Quality Important? • Data deviate from the ‚correct‘ values • Reason (Morgan & Henrion, 1990) – Incompleteness – Disaccord between different data sources – Linguistic uncertainty – Variability – Registration of physical data Gerhard Navratil Workshop GI Product Development

Description of Data Quality • • • Lineage Accuracy Completeness Logical consistency Currency Semantic accuracy (Guptil & Morrison, 1995) Gerhard Navratil Workshop GI Product Development

Storage of Data Quality: Metadata = Data on Data Defined in standards (e. g. , ISO TC 211, 2002) Should be connected with data and contain • Identification • Data quality (without semantics) • Data organization (e. g. , vector - raster) • Information on classification, attributes • Distribution path, contact, liability Gerhard Navratil Workshop GI Product Development

Problems With Metadata • Typically not collected or not updated • User do not understand them (Boin & Hunter, 2007) • No rules for 3 D-data (compare Sargent, Harding & Freeman, 2007) Gerhard Navratil Workshop GI Product Development

What is Quality? „degree to which a set of inherent characteristics fulfils requirements “ (ISO 9000) Requirements depend on application Simple if only one type of application is possible (e. g. , fire arms) What about geographic data? Gerhard Navratil Workshop GI Product Development

Observation and Measurement Observations used to determine properties of the physical world Result measured in different scales – Nominal scale: Names only – Ordinal scale: Includes order – Interval scale: Includes differences – Rational scale: Includes absolute zero Gerhard Navratil Workshop GI Product Development

Total error Difference between a measurement value and the real value. The total error consists of systematic and random error (bias). Gerhard Navratil Workshop GI Product Development

Systematic error (relates to instrument or conditions of measurements) Difference between the mean of measurements and the true value. Systematic errors are determinate errors which affect the accuracy of measurements. They are caused by instrumental-, human-, environmental- or other effects and can be reduced by using control observations and specific observation methods. Instrumental effects e. g. can be reduced by calibration. Gerhard Navratil Workshop GI Product Development

Random error (relates to set of measurements) Difference between a measurement and the mean of measurements. The random error is an indeterminate error (noise) and affects the precision of measurements. Random errors scatter measurements above and below the mean and can be reduced by averaging a set of measurements if the sample set is large enough. Random errors are assumed to be normally distributed. Gerhard Navratil Workshop GI Product Development

Random and Systematic Errors precision: random errors correctness: systematic errors resolution Gerhard Navratil Workshop GI Product Development

Gross error Undetected mistakes that cause a measurement to be very much farther from the mean measurement when compared to other measurements. Gross error Gerhard Navratil Workshop GI Product Development

Accuracy Correctness of a single measurement, calculated from total error. In spatial datasets we can distinguish between positional accuracy and attribute accuracy. Positional accuracy is often divided into vertical and horizontal accuracy (which can differ significantly) and between relative and absolute positional accuracy. Attributes can be measured on four measurement scales: nominal, ordinal, interval and ratio. Error descriptions are different dependent on the used scale. Gerhard Navratil Workshop GI Product Development

Precision Reproducibility of the same measurement value. A statistical measure for precision is the standard deviation. Other common definition for precision: number of digits used to report a measurement, not necessarily related to accuracy! Gerhard Navratil Workshop GI Product Development

Accuracy and precision A B Measurements taken from the same position represented by the center of the circle. Left: successive measurements have similar values, they are precise. But measurements are far from the real value and are therefore inaccurate. Right: Precision is lower but accuracy is higher. Gerhard Navratil Workshop GI Product Development

Vagueness Arises due to poor definition and can be caused by poor documentation fuzziness of objects. Criteria: “Is boundary crisp and well defined? ” “Is the assignment of a particular label to a given zone robust and defensible? ” Gerhard Navratil Workshop GI Product Development

Fitness for Use (1) A function that allows the user to evaluate the fitness of the data for his particular application. Product use decision Difficult: definition of useful output Gerhard Navratil Workshop GI Product Development

Fitness for Use (2) Can I use a data set to solve a specific problem? (Chrisman, 1984) Gerhard Navratil Workshop GI Product Development

Fitness for Use (3) A data set is useful if its use leads to a feasible decision What is the definition of ‚feasible‘? Example: Shop size Gerhard Navratil Workshop GI Product Development

Classification Data are typically classified (organized in classes) – e. g. , land use, language regions, etc. Requires a definition – What is the definition of a forest? (Comber, 2007) Gerhard Navratil Workshop GI Product Development

Uncertainty Reasons for uncertainty? (Fisher, 1999, 2003) • Vagueness (spatial extent of a mountain? ) • Ambiguity (variations in the interpretation of classification rules) • Discord (different classifications – e. g. , land use/land cover) Not considered in standards! (Goodchild, 2007) Gerhard Navratil Workshop GI Product Development

Where are the single buildings? (Förstner, 2007) Gerhard Navratil Workshop GI Product Development

Semantic Loop Gerhard Navratil Workshop GI Product Development

Increasingly Challenging • Data sets for the personal need • Data transfer to colleagues outside of the subject area (Goodchild 2007) Gerhard Navratil Workshop GI Product Development

Problem Traditionally: Map scale describes quality reduction in scale = reduction in quality Typical map scales for different applications Digital maps: Data separated from map scale Idea: Scale introduced during the observation process – carried forward as quality indication? Gerhard Navratil Workshop GI Product Development

Tiered Ontology (1) Ontology: Describes the conceptualization of the world in a particular context (Guarino 1995, Gruber 2005) Ontology for information system: Description of conceptualization and processes (reality and information p. ) Ontology used here: Tiered ontology Gerhard Navratil Workshop GI Product Development

Tiered Ontology (2) • Tier 1: Point observation value from the properties found at a point in space and time: v=p(x, t) • Tier 2: (Physical) Objects Regions with uniformity in property • Tier 3: Social Constructions Constructs relating physical objects to abstract concepts, e. g. , money (Searle 1995) Gerhard Navratil Workshop GI Product Development

Information Process Transforms information obtained at a lower tier to a higher tier Gerhard Navratil Workshop GI Product Development

Observation of physical properties at a point Physical process Links reality (tier 0) to tier 1 Realization is imperfect – Random disturbance – Area observation, not point observation Systematic bias can be eliminated no further consideration Gerhard Navratil Workshop GI Product Development

Object Formation (1) Also called granulation (Zadeh 2002) 2 Steps: – Form boundaries – Summarize properties Meaningful things (Gibson 1986) Gerhard Navratil Workshop GI Product Development

Object Formation (2) Often spatially cohesive solids which move as a single piece In Geography objects do not move uniform properties used to define (overlapping) objects Increases the imperfection of data – summary instead of detailed description Gerhard Navratil Workshop GI Product Development

Boundary Identification Uniform regions Select property and property value, define threshold Object boundary Gerhard Navratil Workshop GI Product Development

Descriptive Summary of properties of the space within the object boundaries Typical functions: Sum, maximum, minimum, average Examples: Weight of a movable object, rainfall on a watershed, maximum elevation in a country (Tomlin 1983, Egenhofer & Frank 1986) Gerhard Navratil Workshop GI Product Development

Classification Mental classification – relating objects to actions, e. g. , using affordances (Gibson 1986, Raubal 2002) Interactions assert conditions Differentiation between objects that fulfill the conditions and those that do not: Distinctions are partially ordered – form a taxonomic lattice (Frank 2006) ‚drinkable‘ is a subtaxon of ‚liquid‘ Gerhard Navratil Workshop GI Product Development

Random Effects in the Observations No perfect observation by physical sensor perturbation of the observation Modeled by Gaussian distribution Effects on – Object formation – Classification – fuzzy membership (Zadeh 1974) Gerhard Navratil Workshop GI Product Development

Effect on Object Formation Statistical error propagation: Gauss‘ Law Summary value: Similar influence Gerhard Navratil Workshop GI Product Development

Scale in Observation Physical observation instruments are small but not infinitely small e. g. , pixel sensor in camera: 5/1000 mm – integrates photons arriving in this region Size effects in the observations are unavoidable scale element Gerhard Navratil Workshop GI Product Development

Physical Point-Like Observation v=p(x, t) is not possible with e covering size and time interval Convolution with a kernel k(e): Formal model for the real observation Convolution with a Gaussian kernel produces an average effect on the signal Gerhard Navratil Workshop GI Product Development

Sampling Theorem Observation density is finite Sampling Danger of artefacts, e. g. , Nyquist Law (sampling twice as frequent as highest signal frequency) Well known for audio signals – applies to all dimensions including sampling in geographic space Gerhard Navratil Workshop GI Product Development

Scale of Observations • Size of smallest objects detected: Extent vs. scale • Effects on uniformity: Variation in property value vs. threshold • Effects on attribute values: Difference max/min vs. average • Effects on object classification: Class distinction by size (large/small building) Gerhard Navratil Workshop GI Product Development

What Follows? Physical observations differ from point observations – Random perturbation of the result – Finite spatial an temporal extent Tier ontology allows to follow the effects – Random variation probability distribution – Finite extent convolution Well designed systems: Influences have comparable size scale Gerhard Navratil Workshop GI Product Development

Influences on Data Quality • Technical limitations (data capture • Legal restrictions (society needs) • User needs Gerhard Navratil Workshop GI Product Development

Technical Limitations Higher quality higher costs Absolute Limit e. g. approximate relative uncretainty of the definition of the unit length ‚meter‘: 2, 5*10 -11 Other examples: Satellite images, GPS Gerhard Navratil Workshop GI Product Development

Legal Restrictions Laws influence data quality by – Influence on data capture processes (e. g. , cadastre, statistics) – Access restrictions Weak constraints (laws can be broken) Examples: Demographic data, data from spatial planning Gerhard Navratil Workshop GI Product Development

User Needs Assumption: Large number of users/important users higher data quality available Indication: Data capture within communities (e. g. , mountain-bikers, hikers, etc. ) Examples: Topographic maps (military), nautical charts Gerhard Navratil Workshop GI Product Development

Legal Aspects for GI Products • Copyright • Data protection • Liability Gerhard Navratil Workshop GI Product Development

Copyright: Definitions • Copyright: Protects the rights of the author Basis: British/American tradition, concentration on economic rights (exploitation rights) • Usage right: Allows possession and (personal) use Gerhard Navratil Workshop GI Product Development

General Definition of Copyright The right to copy; specif. , a property right in an original work of authorship (including literary, musical, dramatic, choreographic, pictorial, graphic, sculptural and architectural works; motion pictures and other audiovisual works; and sound recordings) fixed in any tangible medium of expression, giving the holder the exclusive right to reproduce, adapt, distribute, perform, and display the work. (Black‘s Law Dictionary, 2004) Gerhard Navratil Workshop GI Product Development

Austria: Intellectual Property Right (1) • Definition author: Person creating something unique (text, graphics, design, music, software, etc. ) • Several authors: co-authors • Test of co-authorship: Is the contributor‘s effort an original expression that could quality for copyright protection? (Blacks Law Dictionary, 2004) • Intellectual Property Rights expire – 70 years after the death of the author (USA: 50 years) – 70 years after creation if authors are unclear After that freely usable Gerhard Navratil Workshop GI Product Development

Austria: Intellectual Property Right (2) • Intellectual Property Rights include – Right of reproduction – Right of distribution – Broadcast right – Presentation right – Provision right • Intellectual Property Right ALWAYS stay with the author! • Only parts can be. Gerhard given away Navratil Workshop GI Product Development

Ensembles (Sammelwerke) • Combination of individual contributions to an ensemble are copyright protected if – it constitutes a creation process – the selection of contributions could have been done differently • The ensemble is protected by copyright • But: The individual contributions are still protected by copyright! Gerhard Navratil Workshop GI Product Development

Free Work • Not protected by copyright • Examples: Laws, decrees • Topographic maps of the Austrian surveying authority (BEV) are no free work! Gerhard Navratil Workshop GI Product Development

Exploitation Right (Nutzungsrecht) • Transfer of the rights from the author to a third party • May happen automatically – e. g. , employee writes a text for his company – Important for universities, research centres (exclusive exploitation right? – patents) Gerhard Navratil Workshop GI Product Development

Computer Programs • Work as protected by copyright law if it emerged from an intellectual creation process – Code to access a TXT file is not protected – New method/algorithm to solve a specific type of equation system is protected • Covers source code as well as materials created while developing the program • Employer has unlimited exploitation right if created during office hours Gerhard Navratil Workshop GI Product Development

Copyleft (Slang) • Specific software license model • Allows users to modify or incorporate open-source code into larger programs on the condition that the software containing the source code is publicly distributed without restrictions • Freeware: Software that is made generally available with express or implicit permission for anyone to use, copy, modify, and distribute for any purpose, including financial gain – includes open-source Free refers to usage rather than price! (Blacks Law Dictionary, 2004) Gerhard Navratil Workshop GI Product Development

Databases • Databases are treated like ensembles • No protection if there is only one way to do it, e. g. , – Law database – Probably: addresses, contour lines Gerhard Navratil Workshop GI Product Development

Conclusions Copyright • Protected are display format and selection • No protection if published in form of laws or decrees or if there is no freedom of geometry (boundaries of administrative units) • Problematic: Contour lines Gerhard Navratil Workshop GI Product Development

Personal Data Protection • Right on confidentiality of personal data • Privacy law: Restricts public access to personal information such as tax returns or medical records (Blacks Law Dictionary, 2004) • Informational Privacy: A private person‘s right to choose to determine whether, how, and to what extent information about oneself is communicated to others, esp. sensitive and confidential information (Blacks Law Dictionary, 2004) • Exceptions: – Publicly available data – Aggregated data (if they cannot be re-connected with a person!) Gerhard Navratil Workshop GI Product Development

Definitions • Personal data: Data on persons whose identity is known or can be determined • Indirectly personal: Connection to person cannot be made by legal methods – May be completely impossible – May be possible using illegal means • Sensible data: Special protection (e. g. , racial and ethnic origin, religion, sexual preferences, political or philosophical opinion, health status) Gerhard Navratil Workshop GI Product Development

Liability • The quality or state of being legally obligated or accountable; legal responsibility to another or to society, enforceable by civil remedy or criminal punishment. (Blacks Law Dictionary, 2004) • Austria: Liability if you publicly announce that you are an expert – even if you have no adequate education! (ABGB § 1299) Gerhard Navratil Workshop GI Product Development

Suthradhar vs. Natural Environmental Research Council (1) • British Geological Survey (department of the NERC) works on deep wells (from 1984) • In 1991 -92 the BGS creates a hydro-geologic study on groundwater in Bangladesh • Within the study tests of the water on bacteria and some toxins (aluminium, ferrite, iodine) • 1993 detection of arsenic in the ground water of neighbouring areas • Fault of the experts? Gerhard Navratil Workshop GI Product Development

Suthradhar vs. Natural Environmental Research Council (2) Problem: Original study not for drinking water but how to optimally design tube wells to avoid deterioration Study described as reconnaissance study Results of study applied to creation of wells for drinking water Misuse of data No liability Gerhard Navratil Workshop GI Product Development

When We Have the Data … • How can we access it? • How can we make sure that the communication works? Gerhard Navratil Workshop GI Product Development

Access Methods (1) Communication Channels • Textual • Graphical • Verbal Gerhard Navratil Workshop GI Product Development

Access Methods (2) Limitations of the users? • Language • (Color) Blind • Deaf Gerhard Navratil Workshop GI Product Development

Access Methods (3) Data access • online/offline • Mobile phone/PC • Time of access? Gerhard Navratil Workshop GI Product Development

Information Extraction How to perform each step? • • Collect the data Process the data (e. g. , using a GIS) Create the product Deliver the product through the selected channel Gerhard Navratil Workshop GI Product Development

Feasibility • • • Number of potential customers? Market penetration? Earnings with each customer? Will the system be cost-effective? Return on investment? Gerhard Navratil Workshop GI Product Development

Example: Lunch Menus (1) Problem description • I go to lunch each day and have to make a decision for a specific restaurant • Selection based on menu (changes daily) and personal preference • Menu only visible when I am at the restaurant possible detour if I do not like the menu Solution • GI-product showing me the current menus and the locations of the restaurants Gerhard Navratil Workshop GI Product Development

Example: Lunch Menus (2) Design • User group: people working at the Vienna University of Technology and nearby offices • Access: online via Internet Necessary data • List of restaurant with addresses • Menus of the restaurant (typed in manually) • City map Gerhard Navratil Workshop GI Product Development

Example: Lunch Menus (3) Costs • Simple computer (€ 1000) • Internet access (€ 30/month) • Power supply (€ 30/month) • Data input (2 h/day, € 15/h € 600/month) Number of potential customers • Employees of Vienna University of Technology: • Other employees: Market penetration? • Assessment: 2% Gerhard Navratil Workshop GI Product Development 4000 1000

Example: Lunch Menus (4) Value of the information? • Cost of a lunch € 5 – 14 • Tip: € 0. 5 – 1. 0 • Assessment of value for better decision: € 0. 5 Is the service profitable? • Costs: € 1000 + 660/month • 100 customers/day 2000 queries/month income € 1000/month Not yet included: Programming, marketing, office rent, organization, etc. probably not profitable Gerhard Navratil Workshop GI Product Development

Implementation • Depends on the technology used • Changes rapidly • Should be based on standards and standard protocols (e. g. , GML) • Should consist of independent modules (re -use) Gerhard Navratil Workshop GI Product Development