Tankers to Port

We’ve identified Tankers as a good place to start in this simulation. Let’s build one, along with a Port agent for it to interact with!

Tanker

Our Tanker needs to be able to move towards its destination at a constant velocity. Movement like this can easily be accomplished by adjusting thelng_lat or position property.

// tanker.js
/**
 * Move the tanker to its destination based on its speed.
 * Return its new lng_lat coordinates.
 */
function new_coords(lng_lat, destination) {
  const dx = destination[0] - lng_lat[0];
  const dy = destination[1] - lng_lat[1];

  const vec = hstd.normalizeVector([dx, dy]);

  return [
    lng_lat[0] + vec[0] * state.get("speed"),
    lng_lat[1] + vec[1] * state.get("speed"),
  ];
}

const ll = state.get("lng_lat");
const dest = state.get("destination");

// Move to destination
state.set("lng_lat", new_coords(ll, dest));

On this agent, and many others in the sim, we'll use a published behavior, Lng_Lat to Position, to translate its latitude and longitude into a 3D position.

Note that we need to also set some properties in the globals.json file to use the behavior.

globals.json

{
  "center_ll": [0, 0],
  "scale_ll": 1,
  "seconds_per_step": 60
}

In the finished sim, the tanker.js file behavior also detects whether the agent has reached its destination to begin unloading. You'll also need to logic to initialize.py in order to create the Tanker.

# initialize.py

def behavior(state, context):
    sec_per_step = context.globals()['seconds_per_step']

    agents = []

    agents.append({
      'type': 'tanker',
      'agent_name': 'tanker',
      'lng_lat': [0, 0],
      'destination': [5, 5],
      'behaviors': ['tanker.js', '@hash/ll-to-pos/ll_to_pos.js'],
      'speed': 6.028e-5 * sec_per_step, # convert knots to ll per second
      'scale': [0.5, 1],
      'color': 'green'
    })

    messages = state.get('messages');

    for agent in agents:
      messages.append({
        'to': 'hash',
        'type': 'create_agent',
        'data': agent
      })

    state.set('messages', messages)

Now let’s create a port that the Tanker unloads to.

Port

To give the Port unloading behavior, we'll make use of a published behavior in hIndex called In Flow. If we look at its documentation [here] we can see what properties we need to initialize our Port with. It's not enough just to add this behavior to Port agents, since the Tankers need to cooperate during the unloading process. We'll give them the complementary Out Flow behavior.

# initialize.py

def behavior(state, context):
    sec_per_step = context.globals()['seconds_per_step']

    agents = []

    # Create a tanker
    agents.append({
      'type': 'tanker',
      'agent_name': 'tanker',
      'lng_lat': [0, 0],
      'destination': [5, 5],
      'behaviors': ['tanker.js', '@hash/out-flow/out_flow.js', '@hash/ll-to-pos/ll_to_pos.js'],
      'speed': 6.028e-5 * sec_per_step, # convert knots to ll per second
      'scale': [0.5, 1],
      'color': 'green'
    })

    # Create a port
    agents.append({
        'lng_lat': [5, 5],
        'capacity': 15000,
        'crude': random.uniform(capacity * .2, capacity * .8),
        'behaviors': ['port.js', '@hash/in-flow/in_flow.js', '@hash/ll-to-pos/ll_to_pos.js'],
        'in_flow_property': 'crude',
        'in_nodes': [loc['id'] + ' tanker'],
        'prev_order': 0,
        'flow_rate': 0.1 * sec_per_step, # cubic meters per hour
        'position': None,
        'color': 'purple',
        'scale': [0.5, 0.5]
      })


    # Send create messages for all agents
    messages = state.get('messages');

    for agent in agents:
      messages.append({
        'to': 'hash',
        'type': 'create_agent',
        'data': agent
      })

    state.set('messages', messages)

The final addition to the tanker.js behavior is logic to detect when we're finished unloading and then set a returning property to indicate we should return to the loading location. Test it, and make sure that all the behaviors are working properly! Now let’s build the Refinery that connects to our Port.