Embark SPAC Presentation Deck

Interactive Planning with Data Driven Predictions 60 Second, Interaction-Aware Prediction Horizon Unlocks Optimal Behavior in Challenging Driving Scenarios Intelligent Adaptive Prediction Data driven modeling enables the Embark Driver to predict the actions of other road users Each new observation of a road user improves the system's ability to predict that road user's intent Allowing the Embark Driver to understand if a vehicle intends to yield the right-of-way or close an identified gap Simulating Up To 1200 Scenarios Each Second Each scenario simulates the actions and interactions of the truck and other relevant vehicles in the environment Based on the observed and predicted actions of the other road users The Embark Driver generates a 60 second planning horizon, analyzing relevant actors and available trajectories • ● ● W NORTHERN GENESIS II The plan best fitting multiple criteria including safety, comfort, and route preference is selected and is intended to match a human driver under challenging circumstances Erik Ward PhD Research on Interaction Aware Planning for Self-Driving KTH Royal Institute of Technology Scania Self-Driving Truck Research ● ● ● 1600 1400 1200 1000 800 600 400 200 0 4 Bird's eye view Node: 148, t: 20.6, cst: 49.21 2 0 -1 95 Meters -2 26 24 22 20 18 16 14 2000 1800 1600 1400 1200 1000 800 600 400 200 0 U 10 EGO (-99) o id: 1121446 o id: 1122367 o id: 1121658 10 20 20 Speed profiles 30 30 Travel profiles (selected) 40 40 50 60 13 70 60 80 the

View entire presentation