Startup collaboration within Combient Foundry

Three standout projects from Scania's startup collaborations

1. Simulated environments for training of autonomous vehicles
   
   In this project, Silicon Valley-based startup Applied Intuition provided Scania with powerful tools to simulate and test code for autonomous vehicles more efficiently. The solution significantly reduces the need for physical testing.
   
   "It saves a lot of time and money. Our estimates suggest that autonomous software verification can be developed ten to a hundred times faster - with physical vehicle test only used for late-stage system validation and verification testing," says Jimmy Selling, Partnership Manager at Scania Autonomous Transport Solutions.
   
   
2. Virtual training for electric vehicle service
   
   As electrification accelerates, Scania needs to equip workshops worldwide for battery electric vehicles (BEVs). UK-based startup Digitalnauts developed a virtual training environment that prepares service technicians to safely handle high-voltage batteries - faster and more cost-effectively than traditional methods.
   
   "We no longer need to fly teams around the world or rely on physical mock-ups. That means big savings - my estimate is that it's in the range of EUR 1 million per year," says Patrik Hakanen, Learning Consultant at Scania Academy.
   
   
3. Machine learning speeds up aerodynamic design
   
   Small design changes can significantly affect aerodynamics - but evaluating them usually requires complex, time-consuming calculations. With support from Japanese startup RICOS and U.S.-based PredictiveIQ, Scania's engineers are developing a machine learning tool that delivers highly accurate predictions thousands of times faster than traditional methods.
   
   "We see huge productivity gains in R&D, along with better product performance, improved energy efficiency, and in the end lower CO₂ emissions," says Per Elofsson, an aerodynamics Senior Technical Manager at Scania R&D, who also sees great potential in applying similar methods for other engineering calculations.