Specifications include, but are not limited to: Task 1 – Calibration and setup of AI and ML models/programs • Objective: Setup and calibrate a model/program that uses AI and ML for each of the five projects listed under the “Project Locations” shown above. The objective of the model/program is to develop new traffic signal coordination plans that provide for the orderly and efficient movement of all modes of transportation by providing for smooth and equitable flow of traffic and as outlined in the Requirements and Qualifications, and Task 2 below. • Methodology: The Consultant shall coordinate with the Department to create a model/program that considers the following: a. Signal coordination plans that can be developed quickly, both on the spot, and at a regular interval (such as bi-monthly) for at least three time-of-day periods Monday through Friday (morning, off-peak, afternoon/pm). In addition, at least one time of day period for Saturday and at least one for Sunday. b. Optimize for split time and offsets. The ideal model will also account for cycling sequence changes for better progression, where permissible given safety considerations. c. Limit controller from going into transition. It is assumed that a good model will generally avoid this but the consultant needs to know that split times less than the pedestrian walk + ped clear + yellow + red clear will result in: i. Coordinated Phase: Controller not running coordination ii. Side Street Phase: Controller adjusting the offset for up to multiple cycles after serving a pedestrian Pedestrian volumes typically dictate the Department’s tolerance of lowering the split below the pedestrian minimum time on the side streets. d. Options to provide balanced, one-way, or favored progression schemes e. Cycle length minimum and maximum constraints f. Accommodate minimum split constraints i. Absolute split minimum by phase type (left turn, thru) ii. Pedestrian minimum iii. Coordinated phase minimum