Skip links

Dynamic Route Planning

The project competition takes place until 26/04/2021. There was a LIVE Video workshop on 8 April 2021, please press VIDEO to watch the recording!



Company “Kelių Priežiūra” performs road maintenance tasks on national roads in whole Lithuania (about 21 000 km). These tasks are being performed during both summer and winter. Most of the roads during winter are being treated (about 16 000 km) with slipperiness– reducing materials, such as salts (NaCl, etc.) and sand. Road treatment (spreading) tasks are being performed using more than 300 spreader-trucks and other equipment, that are spread out in 60 maintenance units across Lithuania. 


Over the years the decisions for winter road maintenance tasks were made by on duty employees in the units, and the quality of such decisions were highly dependent on their experience and competence. In 2018, company “Kelių Priežiūra” started using a state-of-the-art road condition forecasting system (Road status information system – RSIS), that produces a new forecast every 1 hour and displays current and future road condition maps.  

RSIS principles:

  • Road network is split into segments with a particular climate characteristic for each, that depends on road structure, natural and anthropogenic environments, and other factors affecting road condition. 
  • Every 1 hour the system collects data from various sources, e.g. road weather stations, numerical weather prediction model, winter road maintenance task information from Purchaser, etc. This data is then processed and a 18h forecast is calculated according to the climate characteristic for each road segment. 
  • The Forecast is available through graphical user interface and application programming interface (API). 
  • When RSIS indicates slippery road sectionse.g. ice on road, it provides a task recommendation. The decision making of whether, where and how many spreaders and/or other equipment is needed is being left for an employee, who makes a call out for salt spreading, road clearing or another task. 
  • During the adverse winter weather at times there is a need to deploy all 300 spreaders and other equipment at once, thus, route planning and resource management becomes an arduous process. Moreover, constantly updating information on road conditions is a crucial factor in decision making, determining the way a task must be performed. 
  • In order for Purchaser to perform the tasks and activities effectively, a need arose for automatic planning and routing of the spreaders and other equipment, that would consider the Forecast and available resources.

Current limitations and factors:

  • Every year before the winter season a set of optimal routes are being made according to the service areas and resources, and will only be adjusted if a significant change in management, road classification, or other area is done.
  • Route correction in operational work according to the weather is performed by on-duty employees, thus, there is a high risk of human factor. 
  • Formation of ice on road is more often patchy than not, therefore, it usually is ineffective to spread slipperiness reducing materials on a whole road.
  • The amount of salt that is needed is usually decided upon by the on-duty employee, who often does not follow RSIS recommendations. 
  • The decision of whether to call-out the spreaders is left for on-duty employees too, instead of a centralised road maintenance centre.
  • The drivers receive the callouts using text or verbal communication, instead of maps or navigation apps. 


Routes shall be generated optimally, being of the shortest length possible and only covering road sections that need to be treated. 

Routes shall take into consideration time restrictions, that are set in the contract of Purchaser and client. 

System shall integrate RSIS information and generate a set of routes according to RSIS Forecast every 1 hour. 

System shall leave the decision whether to choose either Static or Dynamic route for the on-duty employee. 

Drivers shall receive the generated route, task, and other information via an app in a mobile device with Android OS. 

* For more detailed information see the documents in the below section.


Stage 1

Build MVP (minimum viable product)

Stage 2

Participate in contest 

Stage 3

Three winners will be selected

Stage 4

Negotiations with the winners. The contract is signed with the only one winner 


Get ready and Start Pilot!

The project competition takes place until 26/04/2021. There was a LIVE Video workshop on 8 April 2021, please press VIDEO to watch the recording!


Video Workshop

The project competition takes place until 26/04/2021. There was a LIVE Video workshop on 8 April 2021, please press VIDEO to watch the recording!


Skip to content