Data Science
Leveraging all available data assets to help solve complex business problems
As an Intelligent Transport Systems (ITS) solutions provider, Transmax offers STREAMS customers a range of services to enhance the sustainability and performance of customers’ transport networks. This includes services available through the company’s data science team.
Our team of data scientists specialise in identifying, prioritising and undertaking a range of research and analytical activities leveraging all available data assets to help solve complex business problems through the development and application of advanced statistical modelling techniques.
Areas of expertise
Data mining and statistical analysis
Exploratory data analysis to reveal patterns and trends in data. This includes investigating data for inconsistencies and outliers and calculating statistically based measures.
Examples Include:
- Determining confidence measures in traffic related data by providing statistics including significance testing, histograms, boxplots and various other charts
- Estimating vehicle detector health based on speed, volume and occupancy measures and traffic related constraints
- Assessing intersection timing plans by comparing distribution of actual phase times with plan phase times
- Determining typical day statistics for single detectors or groups of detectors so that atypical days can be more easily identified
Visualisations
Providing an accessible way to see and understand trends, outliers, and patterns in seemingly complex traffic related data. These include charts, tables, graphs, maps and dashboards.
Examples include:
- Using Power BI / Tableau to create information rich dashboards
- Interactive web-based visualisations
- Excel related visualisations with macro backend processingining typical day statistics for single detectors or groups of detectors so that atypical days can be more easily identified
Big Data
Analysing and extracting information from traffic related data sets that are seemingly too large to be dealt with in a traditional sense. This typically involves cloud computing on Amazon Web Services (AWS).
Examples include:
- Aggregating large traffic related data sets to various time intervals based on 20-second data
- Visualising large aggregated data sets in various formats using cloud-based tools
Machine Learning
Creating software that utilises data, and is capable of learning from past information, thereby offering actionable predictions.
Examples include:
- Cluster analysis to identify time-of-day traffic patterns
- Image recognition to determine speed limit signs by location using Google Maps
- Predict SVO time series variables based on historical data