Traceability

Data tends to “travel” over time and distance; this data source refers to the ability to analyse and discover value in data, generated by tracing people and items across the globe. Typical questions to be answered: 

 
  • Do your customers or users generate traceable data, do these data “travel”?
  • Are your products shipped? Is logistic excellence a KPI for you?
  • Are weather, traffic and/or other environmental parameters impacting your business?
  • Do you store data regarding time, distance and/or location-series that require real time monitoring and follow-up?

Data cases:

  • Track & trace monitoring systems.
  • Point of sale location analysis.
  • Remote monitoring systems for healthcare, e.g. ambulatory patients and the elderly at home.
  • Data driven crowd and trajectory analysis, e.g. mass event security, city marketing, retail marketing

Data story (combined with Behavibility©)

"In collaboration with Mobistar, the city of Antwerp wanted to explore the possibilities to monitor crowds during a series of public events: e.g. a fix location with a visiting period of 7 weeks or a sports event with very high density crowds during one day. Cropland analyzed the mobile data sets and proposed a methodology to separate local contacts, independent of the event, from visitors’ contacts. This approach allowed computing the real number of daily visitors and how they behaved during their journey. For the sports event, the crowd monitor tool was offered as a real time dashboard for local security"
Press release:

Belga NL:http://www.belga.be/nl/press-release/details-53879/?langpr=NL

Belga FR: http://www.belga.be/fr/press-release/details-53879/?langpr=FR

Crowd Monitoring Tool in real time - Antwerp 10 Miles, April 17, 2016 Data are altered for security reasons

Crowd Monitoring Tool in real time - Antwerp 10 Miles, April 17, 2016

Data are altered for security reasons

Readability

The data science for textual information; it refers to the ability to automatically recognize, process, analyze and predict the underlying meaning, message and/or sentiment in unstructured textual data. Typical questions to be answered: 

  • Is your business document/text driven?
  • Is document processing a key element of your daily operations?
  • Can the content be interpreted?
  • How can predictive value of textual information be unlocked?

Data cases: 

  • Document categorisation, tagging, enrichment and summarization
  • Sentiment analysis-based workflows
  • Data quality detection in text
  • Automated message triage and dispatching systems

Data story

"A software developer for legal services required a more automated process for document content handling. Cropland introduced a big data text mining methodology based on specific terminology and key words that could be found in the documents. Instead of the full ‘start-from-scratch’ manual labeling activities in the past, Cropland was able to develop an algorithm, which screened contents and auto-proposed labeling only to be checked and/or corrected by the companies’ editors."

Connectability

The Internet of Things or a world where data emerges from connected devices, utilities and sensors. It allows to find and predict patterns within the multiple layers of machine communication. Typical questions to be answered: 

  • Are your products connected to the Internet or mobile network?
  • Do your consumers / users track their activity?
  • Are your production plants or finished products equipped with connected sensors?
  • Are your employees digitally connected?

Data cases:

  • Utility management of buildings and/or premises
  • Learning domotica at home
  • Connected and predictive healthcare based on wearable and smart device analysis
  • Security and safety prevention
  • Stress & burn-out detection

Data story

"A customer in charge of logistics had only an overall look on the delivery accuracy of their goods. As they were working with different third parties in parcel delivery, they wanted to have a view on recurrent anomalies with strong negative impact at end customer level. Cropland analyzed the data via pattern mining and revealed different delivery scores according the day of the week. On top, daily deliveries with small quantity parcels at the same addresses showed a clear inefficiency in the way the end customer was serviced. Cropland conducted an optimization exercise and showed the need for an internal reorganization of warehouses."

Behavibility©

Cropland introduces 'behavibility', the data science that analyses entity behavior; it refers to the ability to predict and even, prescribe the behavior of customers as being individual consumers, governments or organisations. Typical questions to be answered: 

  • Who are your customers internal & external, people and organisations?
  • How do you differentiate?  What’s their ‘consumption’ behavior?
  • How does the purchasing decision process work?
  • Can we pro-actively involve them in our offering?
  • What’s their retention risk and should we act on churn, or not?

Data cases:

  • Customer behavior analysis including omni-channel purchase insights, purchase drivers, churn & loyalty analysis, fraud detection, credit scoring, etc.
  • Recommendation systems for optimizing product offerings and cross-sell opportunities.
  • Customer segmentation and profiling including lead generation & qualification.
  • HR related profiling such as retention analyses, profile search optimization, job–employee matching, etc.

Data story

"A manufacturer of raw materials in the construction sector wanted to generate a list with leads based on the end customers’ sector classification code. As they were unsuccessful in the past to optimize this lead generation, they asked Cropland for a new approach. Text-mining algorithms were used to explore different open data sources in order to match it with key words indicating a probability score of a fit/non-fit lead according the NACE-code. The success rate of outbound marketing activities based on the Cropland’s prospect list was significantly higher than other former attempts to find the right potential customer."

Data story (in combination with Traceability)

"In collaboration with Mobistar, the city of Antwerp wanted to explore the possibilities to monitor crowds during a series of public events: e.g. a fix location with a visiting period of 7 weeks or a sports event with very high density crowds during one day. Cropland analyzed the mobile data sets and proposed a methodology to separate local contacts, independent of the event, from visitors’ contacts. This approach allowed computing the real number of daily visitors and how they behaved during their journey. For the sports event, the crowd monitor tool was offered as a real time dashboard for local security"
Press release:

Belga NL:http://www.belga.be/nl/press-release/details-53879/?langpr=NL

Belga FR: http://www.belga.be/fr/press-release/details-53879/?langpr=FR

Predictive maintenance

The data science that analyses machine behavior; it refers to the ability to predict and even, prescribe the behavior of machines as being an intelligent network. Typical questions to be answered: 

  • What patterns do I see before fall out?
  • How can we develop an early warning system?
  • Can we create feedback to the system and make it 'learning'?
  • Can we connect it to other systems?

Data cases:

  • M2M data analysis and predictive modeling
  • Pattern mining for power grids
  • Spare part management
  • The 'learning' machine