2d:null 2e:["$","title",null,{"children":"Employees Transportation | NB.AI Solutions Team | Observable"}] 33:Tb90,### Background In this notebook, we will apply a combination of NextBillion's components, namely the Optimization, Distance Matrix and Web Map SDK, to provide one possible solution to the Employee Transportation Use-Case. In this Use-Case the employees satisfaction is the main criteria when evaluating the quality of the transportation service. As such three main requirements will need to be satisfied (as expressed by employees themselves): * _"I want to spend minimal time in the minivan to go to work."_ * _"I do not want to wait too long for the minivan to arrive at my place."_ * _"I do not want to reach the office too much in advance before my actual work starts."_
### Problem Statement The employer, with one central office (factory) proposed to his 200 employees to pick them up every morning to take them to the office. The employees are spread across the city or suburb. Up to 50 minivans, starting from the office, are available to complete the employees transportation. The drivers shift are from 06:00 to 09:00am. Each minivan has a capacity of 5, driver excluded. The key criteria of a successful service is employee satisfaction, as such the following needs to be taken into account when planning: * An accurate pickup time should be provide to each employee to minimize the wait time of the minivan. * Each employee should be picked-up onward (starting with the furthest employee) to the office to minimize time spend in transport. * The minivans should reach ideally the office between 08:45 and 09:00am, just before the employees shift starts, to minimize wait time at the factory.
### Approach The approach we will follow to propose one solution to this problem is described below: * Taking into account all employees locations, available minivans, and constraints (capacity, driver shift), compute the optimal route for each driver to pickup up to 5 employees. This will lead to groups of employees assigned for pickup to different minivans. * Compute for each group of employees a Distance Matrix from office location to each employee location to find the furthest employee location. * For each minivan, set the driver start location to the furthest employee location and end location to the office. * Add an artificial job at the destination (office) with a window time constraint between 08:45am and 09:00am. * Run Optimization within each group with these new constraints. * Render results.
In this tutorial, we'll guide you through the steps that you can take to build, solve, and visualize complex multi-vehicle routing problem applied to this particular Use-Case.

Refer to the API documentation at this page for more supporting details: [API doc](https://docs.nextbillion.ai/docs/navigation/api/optimization-mvrp)

Note: this demo has the possibility of generating a large number of MVRP transactions against the key being used and should only be used with evaluation/demo API keys. 34:T5c0,tempMVRPProblem = { if (run_optimization && inputData_a) { let location = ""; let loc_index = 0; let shipments = []; let jobs = []; let vehicles = []; let jobs_index = 0; randomVRPInputs.pointsArray.forEach( p => { let job = { id: p.id, description: p.business, pickup: [1], location_index: loc_index++, service: parseInt(inputData_a.serviceTime) } jobs.push(job); if (jobs_index != 0 && jobs.length <= randomVRPInputs.pointsArray.length){ location += '|'; } location += `${p.pickup_latitude},${p.pickup_longitude}`; jobs_index++; }); let vehicles_index = 0; randomVRPInputs.vehicleArray.forEach( f => { if (vehicles_index < randomVRPInputs.vehicleArray.length){ location += '|'; } location += `${f.latitude},${f.longitude}`; //location += panel_random.city.centroid; let vehicle = { id: f.id, description: f.vin, start_index: loc_index, end_index: loc_index++, time_window: [Date.parse(inputData_a.shiftStart)/1000, Date.parse(inputData_a.shiftEnd)/1000], capacity: [ parseInt(inputData_a.capacity) ] } vehicles.push(vehicle); vehicles_index++; }); let locations = { id:1, location: location } return {locations: locations, location_index: loc_index, jobs: jobs, vehicles: vehicles}; } } 35:T8d2,selectedMVRPProblem = { if (run_optimization && inputData_a) { let location = ""; let loc_index = 0; let shipments = []; let jobs = []; let vehicles = []; let jobs_index = 0; selectedRoute[0].steps.forEach( step => { let job = {} if (step.type == "job"){ job = { id: step.id, description: step.business, pickup: [1], location_index: loc_index++ //service: parseInt(inputData_a.serviceTime) } jobs.push(job); if (jobs_index != 0 && jobs.length <= (selectedRoute[0].steps.length - 2)){ location += '|'; } location += step.location[0] + "," + step.location[1] ; jobs_index++; } }); let tomorrowStart = new Date(); tomorrowStart.setDate(new Date().getDate()+1); tomorrowStart.setHours(8); tomorrowStart.setMinutes(45); tomorrowStart.setSeconds(0); tomorrowStart.setMilliseconds(0); let tomorrowEnd = new Date(); tomorrowEnd.setDate(new Date().getDate()+1); tomorrowEnd.setHours(9); tomorrowEnd.setMinutes(0); tomorrowEnd.setSeconds(0); tomorrowEnd.setMilliseconds(0); var startEpoch = tomorrowStart.getTime()/1000.0; var endEpoch = tomorrowEnd.getTime()/1000.0; console.log("#################"); console.log(startEpoch); console.log(endEpoch); let fakeJob = { id: 9999, description: "Job at office with time window", pickup: [0], location_index: loc_index++, time_windows: [[startEpoch,endEpoch]] } jobs.push(fakeJob); location += "|" + officeLocation ; jobs_index++; // vehicle start & end location location += "|" + furthestEmployee.location[0] + "," + furthestEmployee.location[1]; location += "|" + officeLocation ; let vehicle = { id: 1, description: routeSelector.properties.vin, start_index: loc_index++, end_index: loc_index++, capacity: [ parseInt(inputData_a.capacity) ] } vehicles.push(vehicle); //console.log(location); let locations = { id:1, location: location } return {locations: locations, jobs: jobs, vehicles: vehicles}; } }36:T5ad,inputs = { let tomorrowStart = new Date(); tomorrowStart.setDate(new Date().getDate()+1); tomorrowStart.setHours(8); tomorrowStart.setMinutes(0); tomorrowStart.setSeconds(0); tomorrowStart.setMilliseconds(0); tomorrowStart.toLocaleDateString(); let tomorrowEnd = new Date(); tomorrowEnd.setDate(new Date().getDate()+1); tomorrowEnd.setHours(11); tomorrowEnd.setMinutes(0); tomorrowEnd.setSeconds(0); tomorrowEnd.setMilliseconds(0); tomorrowStart.toLocaleDateString(); const inputs = { city: Inputs.select(cities, {label: "Select city", format: x => x.name, value: cities.find(t => t.name === "Paris")}), jobs: Inputs.select(jobs,{label: "Number of Employees",value: "25"}), vehicles: Inputs.select(vehicles,{label: "Number of vehicles",value: "10"}), shiftStart: Inputs.datetime({label: "Shift Start", value: tomorrowStart.toISOString().slice(0, -1)}), shiftEnd: Inputs.datetime({label: "Shift End", value: tomorrowEnd.toISOString().slice(0, -1)}), useCase: Inputs.select(useCases, {label: "Use Case", format: x => x.name, value: useCases.find(t => t.name === "service")}), objectiveFunction: Inputs.select(objectives, {label: "Objective Function", format: x => x.name, value: objectives.find(t => t.name === "Minimize Time")}), serviceTime: Inputs.text({label: "Service Time at Each Stop", value: "30"}), capacity: Inputs.text({label: "Capacity", value: "5"}) }; return inputs; }37:T4c6,routesGeoJson = { let routesGeoJson = { type: 'FeatureCollection', features: [] }; // we create a geojson object to send to our mapview elements // It will be a feature collection of all the route polylines checkStatus.result.routes.forEach((rte,idx) => { //console.log(rte); let feature = { type: 'Feature', properties: { id: rte.vehicle, vin: rte.description }, geometry: { coordinates: nextbillion.utils.polyline.decode(rte.geometry,5) .map((c) => c.reverse()), type: 'LineString' } } routesGeoJson.features.push(feature); rte.steps.forEach((st,stIdx) => { let stop = { type: 'Feature', properties: { id: rte.vehicle, vin: rte.description, description: st.description, stop_idx: stIdx, arrival: st.arrival, type: st.type, pointType: ((st.type == "start"|st.type == "end")?16:255) }, geometry: { coordinates: [st.location[1],st.location[0]], type: 'Point' } } routesGeoJson.features.push(stop); }) }); console.log(routesGeoJson); return routesGeoJson; }38:T4cd,routesGeoJson2 = { let routesGeoJson2 = { type: 'FeatureCollection', features: [] }; // we create a geojson object to send to our mapview elements // It will be a feature collection of all the route polylines checkStatus2.result.routes.forEach((rte,idx) => { //console.log(rte); let feature = { type: 'Feature', properties: { id: rte.vehicle, vin: rte.description }, geometry: { coordinates: nextbillion.utils.polyline.decode(rte.geometry,5) .map((c) => c.reverse()), type: 'LineString' } } routesGeoJson2.features.push(feature); rte.steps.forEach((st,stIdx) => { let stop = { type: 'Feature', properties: { id: rte.vehicle, vin: rte.description, description: st.description, stop_idx: stIdx, arrival: st.arrival, type: st.type, pointType: ((st.type == "start"|st.type == "end")?16:255) }, geometry: { coordinates: [st.location[1],st.location[0]], type: 'Point' } } routesGeoJson2.features.push(stop); }) }); console.log(routesGeoJson2); return routesGeoJson2; }39:[] 31:{"id":"9f05f36c822642e4","slug":"employees-transportation","title":"Employees Transportation","thumbnail":"4ca482cbf408fca333a18779c970bcd1549030ccaec26a0aab329153588b69f0","access_level":"public","create_time":"$D2023-01-16T08:38:34.720Z","update_time":"$D2023-02-01T09:41:16.878Z","view_time":null,"first_public_time":"$D2023-01-25T08:39:50.924Z","has_unpublished_edits":null,"fork":true,"stars":0,"starred":false,"owner":{"type":"team","id":"05baed136f9134c1","login":"nbai","name":"NB.AI Solutions Team","avatar_url":"https://avatars-next.observableusercontent.com/7a25d1b5deb9c5290b12e4e0914aacdbd4196bb5c3a94c63a0fda0fe706d28b2","tier":"starter_2024"},"version":2555,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"014ccaba67132a7a","login":"yvannbai","name":"yvanNBAI","avatar_url":"https://avatars-next.observableusercontent.com/a43e99fc0d5f61a6d2e1ba9995c0d5f9e0c33e87df89f044ce85e127d00df06c","tier":null}],"members":[{"type":"team","id":"05baed136f9134c1","login":"nbai","name":"NB.AI Solutions Team","avatar_url":"https://avatars-next.observableusercontent.com/7a25d1b5deb9c5290b12e4e0914aacdbd4196bb5c3a94c63a0fda0fe706d28b2","tier":"starter_2024","role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":4,"collection":null,"type":"notebook","body":{"title":"Employees Transportation","cells":[{"id":0,"value":"![text](https://nextbillion.ai/wp-content/uploads/2022/01/logo-nb-colored.png)\n# Employees Transportation \n# Employees Satisfaction","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2518,"value":"---","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2517,"value":"toc(\"h2, h3, h4\")","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2516,"value":"---\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":508,"value":"$33","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":1189,"value":"viewof apiKey = secret2(\"apiKey\", {\n title: \"NextBillion.ai API Key\",\n submit: \"Set\",\n description: \"Contact support@nextbillion.ai to get your API key\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":525,"value":"### Random Data Generation\nIn this section we will set some initial parameters that are used to formulate the inputs to the MVRP solver. The `NB.ai API KEY` is necessary to interact with the MVRP services and it should be entered first to activate the tutorial tasks.\n

\nSelect City Selection, Number of Employees, Number of Vehicles. You can leave the default. Note for this specific tutorial, we limit the total number of employees to 25 maximum, but it does not change the overall logic. \n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":1234,"value":"---\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2024,"value":"viewof panel_random = Inputs.form(inputs, {template: v => htl.html`\n

\n ${v.city}\n ${v.jobs}\n ${v.vehicles}\n

\n\n`})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1566,"value":"{\n const button_random = html` `;\n button_random.onclick = () => {\n mutable inputData_a = panel_random;\n button_random.disabled = true; \n }\n return html`${button_random}`;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1236,"value":"---\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":534,"value":"{\n return html`

${md`### Random Data Visualization\nLet's start with the retrieval of random vehicle and employees data for the city that was selected. The random data will contain the requested number of jobs and each will consist of a pickup location. Feel free to browse the object to see the pickup content. Green points are vehicles and red points are pickup jobs.`}

`;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":964,"value":"{ \n if (!apiKey && randomVRPInputs) {\n\n\n \n return geoviewer(randomVRPInputs.vehGeoJson, {width:750,height:450,col:'#8058a0',fillOpacity:.9,lineWidth:5,style:'positron'});\n } else {\n if (randomVRPInputs) {\n return geoviewer(randomVRPInputs.vehGeoJson,{width:\"100%\",height:450,col:'#8058a0',fillOpacity:.9,lineWidth:5,style:'nbai',apiKey:apiKey});\n }\n }\n\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1912,"value":"tempRandomVRPInputs = {\n let retval = null;\n try {\n let retval = axios({\n method: \"GET\",\n url: `https://m4aqpzp5ah6n7ilzunwyo5ma2u0ezqcc.lambda-url.us-east-2.on.aws/?region=${inputData_a.city.value}&number=${inputData_a.jobs}&vehicles=${inputData_a.vehicles}&type=${inputData_a.useCase.value ==='delivery'?'darp':'service'}`,\n headers: {\n } \n }).then(res=>res.data); \n return retval;\n } catch (err) {\n }\n}\n","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2490,"value":"### Problem Constraints\n\nHere we are setting a few constrains including the minivan drivers shift, vehicle maximum capacity and objective function to use.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2122,"value":"// Transform JSON output from Lambda mock to ensure all vehicles starts from same location\n// The vehicles locations are set to city center\n\nrandomVRPInputs={\n let returnVal = null;\n let vehiclesLoc = (panel_random.city.centroid).split(\",\");\n\n returnVal = async () => {\n let tmp = await tempRandomVRPInputs;\n \n tmp.vehicleArray.forEach( vehicle => {\n vehicle.latitude = vehiclesLoc[0];\n vehicle.longitude = vehiclesLoc[1];\n });\n \n for (let i = 0; i < tmp.vehicleArray.length; i++) {\n tmp.vehGeoJson.features[i].geometry.coordinates[0] = vehiclesLoc[1];\n tmp.vehGeoJson.features[i].geometry.coordinates[1] = vehiclesLoc[0];\n }\n \n //console.log(tmp);\n return tmp;\n };\n\n return returnVal();\n}\n","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2038,"value":"viewof panel_mvrp = Inputs.form(inputs, {template: v => htl.html`\n

\n ${v.shiftStart}\n ${v.shiftEnd}\n ${v.objectiveFunction}\n ${v.capacity}\n

\n\n`})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2042,"value":"{\n const button_mvrp = html` `;\n button_mvrp.onclick = () => {\n mutable run_optimization = true;\n mutable inputData_a = panel_mvrp;\n button_mvrp.disabled = true; \n }\n return html`${button_mvrp}`;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":206,"value":"---\n### Build the Problem Inputs\nWe start to build the needed JSON objects that will be submitted to the MVRP endpoint. In this section, we are building the `jobs` which each consist of employees pickup locations in the city region. We also create the list of locations that will be referenced by index. Browse the temporary object that will be used further down the page to create the final input object. You can also expand the block to see the Javascript code that created the temporary object.\n\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":190,"value":"$34","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":594,"value":"---\n### Submit the Problem to the MVRP API Endpoint\nWe are now to the point where we can put together the full MVRP request containing all of our pickup locations, vehicles, and options for the solution.\n

\nThe MVRP endpoint is a non-blocking/asynchronous endpoint. We will receive back a reference ID for the optimization job and a status message as seen below. We've highlighted the reference ID in green.\n

\nNote: if you see an error in this section, more than likely you have an invalid API Key set in the input box above. \n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":221,"value":"mvrpID = {\n\n const mvrpRequest = {\n description: `MVRP Tutorial ${new Date().toISOString()}`,\n locations: tempMVRPProblem.locations,\n vehicles: tempMVRPProblem.vehicles,\n jobs: tempMVRPProblem.jobs,\n options: {\n objective: {\n travel_cost: inputs.objectiveFunction.value.value,\n minimise_num_depots: false\n }\n }\n };\nconsole.log(JSON.stringify(mvrpRequest));\nlet resp = axios.post(`https://api.nextbillion.io/optimise-mvrp?key=${apiKey}`, mvrpRequest)\n .then(response => response.data )\n .catch(response => response.data)\n return resp;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":348,"value":"viewof jobID = htl.html`
MVRP Request ID: ${mvrpID?mvrpID.id:mvrpID}
`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":601,"value":"---\n### Retrieve MVRP Result\nOnce we have the reference ID for the MVRP request, we can poll that MVRP status endpoint to get the solution. In this tutorial, we are waiting 10 seconds before polling for the solution. When you see a status message of `Ok` you can browse the result to see the optimized routing solution.\n

\nNote: if you see an error in this section, more than likely you have an invalid API Key set in the input box above or the API key is not valid for the city region selected.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":720,"value":"html`
If you get a status message indicating Job still processing just wait a few seconds and press this button.

${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":242,"value":"checkStatus = {\n mvrpID;\n //await visibility();\n requestResult;\n yield Promises.delay(5000).then(() => {\n let mvrpResult = axios({\n method: \"GET\",\n url: `https://api.nextbillion.io/optimise-mvrp/result?id=${mvrpID.id}&key=${apiKey}`,\n headers: {\n } \n }).then(res=>res.data)\n .catch(error => {\n let mvrpResult = {\n status: 'Error', \n error: JSON.stringify(error)\n }\n });\n return mvrpResult;\n })\n}\n","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":631,"value":"kpiStats = {\n let deliveryCount = 0;\n checkStatus.result.routes.forEach((rte, rteidx) => {\n rte.steps.forEach((step, index) => {\n if (step.id && step.id != 99) {\n ++deliveryCount;\n }\n });\n });\n return {\n totalDistance: `${(checkStatus.result.summary.distance/1000).toFixed(1)} km`,\n totalTime: `${(checkStatus.result.summary.duration/60).toFixed(1)} mn`,\n nbrRoutes: checkStatus.result.summary.routes,\n avgStops: (deliveryCount/checkStatus.result.summary.routes).toFixed(1)\n }\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2414,"value":"### KPIs for all minivans","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":645,"value":"| KPI | Value |\n| ------ | -------- |\n| Total Distance | ${kpiStats.totalDistance} |\n| Total Time | ${kpiStats.totalTime} |\n| Number of Routes | ${kpiStats.nbrRoutes} |\n| Average Stops Per Vehicle | ${kpiStats.avgStops}|","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":605,"value":"---\n### Render Solution on a Map\nThis maps shows groups of employees assigned to each minivan. \n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":1261,"value":"viewof routeSelector = Inputs.select(routesGeoJson.features.filter(f => f.geometry.type === 'LineString'), {multiple: false, label: \"Select route to view\", format: x => x.properties.vin, value: t => t.properties.vin})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1676,"value":"green marker is the vehicle start location and red markers are stop locations.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":997,"value":"{\n return geoviewer(routesFiltered, {width:\"100%\",height:\"550\",col:'#8058a0',fillOpacity:.9,lineWidth:3,style:'nbai',apiKey:apiKey});\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1862,"value":"Inputs.table(selectedRoute[0].steps, \n {\n columns: [\n \"type\",\"arrival\",\"load\",\"description\",\"duration\"],\n header: {\n type: \"Task\",\n arrival: \"Arrival Time\",\n load: \"Nbr Loaded\",\n description: \"Account\",\n duration: \"Duration\"\n },\n width: {\n type: 80,\n arrival: 120,\n load: 60,\n duration: 100\n },\n format: {\n type: x => x.toUpperCase(),\n arrival: x => (new Date(x*1000).toLocaleTimeString([],{hour:'2-digit',minute:'2-digit',hour12: false})),\n duration: x => (x/60).toFixed(1) + ' mn'\n },\n rows: 10,\n maxWidth: 600,\n multiple: false,\n required: false\n })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1898,"value":"selectedRoute = checkStatus.result.routes.filter(f => f.description === routeSelector.properties.vin)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2359,"value":"selectedRoute2 = checkStatus2.result.routes.filter(f => f.description === routeSelector2.properties.vin)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2164,"value":"---\n### Optimize each driver journey to meet employees requirements\n\nHere we want to set the furthest employee from the office to be the start location of the driver journey. This will ensure the employees are picked in the appropriate order from the furthest to the closest, thus to minimize the time each employee spent in the minivan. ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2173,"value":"### Distance Matrix Computation\n\nFor each optimized group of employees created by the optimizer (all the employees to be picked up by a given driver), we will assess who is the furthest employee. As such we compute a Distance Matrix where the origin is the office and the destinations are the employee locations.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2191,"value":"distanceMatrix = {\n let destinations = \"\";\n let officeLL = panel_random.city.centroid.split(\",\");\n let origins = officeLocation;\n let job_index = 0;\n selectedRoute[0].steps.forEach(step => {\n if (step.type == \"job\"){\n if (job_index != 0){\n destinations += \"|\";\n }\n destinations += step.location[0] + \",\" + step.location[1];\n job_index++;\n \n }\n });\n\n let dmRetVal = null;\n try {\n let dmRetVal = axios({\n method: \"GET\",\n url: `https://api.nextbillion.io/distancematrix/json-concise?key=${apiKey}&mode=4w&origins=${origins}&destinations=${destinations}`,\n headers: {\n 'Content-Type': 'application/json'\n }\n }).then(res=>res.data); \n return dmRetVal;\n } catch (err) {\n }\n \n //console.log(origin);\n //console.log(destinations);\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2214,"value":"### Furthest Employee\n\nHere we apply some basic logic by parsing the distance matrix output to pick the employee for which the distance is the furthest from the office.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2218,"value":"furthestEmployee ={\n let step_index = 0;\n let highest_dist_index = 0;\n let highest_dist = 0;\n distanceMatrix.rows[0].forEach((el,idx) => {\n if (el[1] > highest_dist) {\n highest_dist = el[1];\n highest_dist_index = idx;\n }\n });\n \n step_index = highest_dist_index + 1;\n return selectedRoute[0].steps[step_index];\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2418,"value":"---\n### Build the Problem Inputs for each Employee group\nNow we build the needed JSON objects that will be submitted to the MVRP endpoint for each employee group. We are as before building the `jobs` which each consist of employees pickup locations in the city region with the main difference that we apply the optimizer to a subset i.e one single employee group. In addition we set new constraints:\n* the driver start location is the furthest employee location to ensure the furthest employee is picked up first.\n* an artificial job (with no pickup) is created at the office location with a pickup time window constraint set from 08:45 to 09:00 am.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2287,"value":"$35","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2289,"value":"mvrpID2 = {\n\n const selecteMvrpRequest = {\n description: `MVRP Tutorial ${new Date().toISOString()}`,\n locations: selectedMVRPProblem.locations,\n vehicles: selectedMVRPProblem.vehicles,\n jobs: selectedMVRPProblem.jobs,\n options: {\n objective: {\n travel_cost: inputs.objectiveFunction.value.value,\n minimise_num_depots: false\n }\n }\n };\nconsole.log(JSON.stringify(selecteMvrpRequest));\nlet resp = axios.post(`https://api.nextbillion.io/optimise-mvrp?key=${apiKey}`, selecteMvrpRequest)\n .then(response => response.data )\n .catch(response => response.data)\n return resp;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2292,"value":"viewof jobID2 = htl.html`
Selected MVRP Request ID: ${mvrpID2?mvrpID2.id:mvrpID2}
`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2294,"value":"---\n### Retrieve MVRP Result\nOnce we have the reference ID for the MVRP request, we can poll that MVRP status endpoint to get the solution. In this tutorial, we are waiting 10 seconds before polling for the solution. When you see a status message of `Ok` you can browse the result to see the optimized routing solution.\n

\nNote: if you see an error in this section, more than likely you have an invalid API Key set in the input box above or the API key is not valid for the city region selected.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2313,"value":"html`
If you get a status message indicating Job still processing just wait a few seconds and press this button.

${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2341,"value":"checkStatus2 = {\n mvrpID2;\n //await visibility();\n requestResult;\n yield Promises.delay(5000).then(() => {\n let mvrpResult2 = axios({\n method: \"GET\",\n url: `https://api.nextbillion.io/optimise-mvrp/result?id=${mvrpID2.id}&key=${apiKey}`,\n headers: {\n } \n }).then(res=>res.data)\n .catch(error => {\n let mvrpResult = {\n status: 'Error', \n error: JSON.stringify(error)\n }\n });\n return mvrpResult2;\n })\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2344,"value":"kpiStats2 = {\n let deliveryCount = 0;\n checkStatus2.result.routes.forEach((rte, rteidx) => {\n rte.steps.forEach((step, index) => {\n if (step.id && step.id != 99) {\n ++deliveryCount;\n }\n });\n });\n return {\n totalDistance: `${(checkStatus2.result.summary.distance/1000).toFixed(1)} km`,\n totalTime: `${(checkStatus2.result.summary.duration/60).toFixed(1)} mn`,\n nbrRoutes: checkStatus2.result.summary.routes,\n avgStops: (deliveryCount/checkStatus2.result.summary.routes).toFixed(1)\n }\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2346,"value":"---\n### Render Solution on a Map","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":2348,"value":"viewof routeSelector2 = Inputs.select(routesGeoJson2.features.filter(f => f.geometry.type === 'LineString'), {multiple: false, label: \"Select route to view\", format: x => x.properties.vin, value: t => t.properties.vin})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2354,"value":"{\n return geoviewer(routesFiltered2, {width:\"100%\",height:\"550\",col:'#8058a0',fillOpacity:.9,lineWidth:3,style:'nbai',apiKey:apiKey});\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2361,"value":"Inputs.table(selectedRoute2[0].steps, \n {\n columns: [\n \"type\",\"arrival\",\"load\",\"description\",\"duration\"],\n header: {\n type: \"Task\",\n arrival: \"Arrival Time\",\n load: \"Nbr Loaded\",\n description: \"Account\",\n duration: \"Duration\"\n },\n width: {\n type: 80,\n arrival: 100,\n load: 80,\n description: 140,\n duration: 100\n },\n format: {\n type: x => x.toUpperCase(),\n arrival: x => (new Date(x*1000).toLocaleTimeString([],{hour:'2-digit',minute:'2-digit',hour12: false})),\n duration: x => (x/60).toFixed(1) + ' mn'\n },\n rows: 10,\n maxWidth: 600,\n multiple: false,\n required: false\n })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2467,"value":"viewof driverDeparture = htl.html`
For the driver to arrive at the first pickup location (the furthest employee) at time indicated in the table above, he will need to start from the office at ${driverDepartureTime}
`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2443,"value":"### Conclusion\n\nUsing a multi-steps approach where the optimizer is first used as a pre-processing step to create employees groups, running the distance matrix inside each group to assess who is the furthest employee and finally running the optimization again inside each group, we manage to achieve the set requirements as highlighted in the table above, which ultimately should lead to employee satisfaction.\n\nBy following this multi-step approach we were able to:\n* Provide a time to each employee for the pickup.\n* Minimize the time spend in the minivan.\n* Arrive a few minutes before the actual shift starts.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":1000,"value":"\n\n\n\n---\n### Additional Tools \nIf you want to visualize the optimized routing solution along with the sequencing of the stops, the NB.ai Optimization Tester tool is available. \n\nNB.AI MVRP Playground App (should automatically load the current solution)\n- Select the `SOLUTION` tab\n- Enter the same NB.ai API key used for the MVRP solution `${apiKey}`\n- Paste the MVRP Request ID into the input box `${mvrpID.id}`\n- Press the `Load` button","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":341,"value":"### Other Dependencies","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":1139,"value":"import {nextbillion,cities,secret2} from \"@yvannbai/nextbillion-ai-helpers\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":334,"value":"axios = {\n let axios = await require('axios');\n return axios;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":352,"value":"styling = html`\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":849,"value":"geoviewer = require(\"https://d12qcqjlhp2ahm.cloudfront.net/nb_geoverview2.js\").then((f) => f.view)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1613,"value":"html`${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1005,"value":"viewof requestResult = Inputs.button(\"Request Result\")","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1686,"value":"mutable inputData_a = null","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2051,"value":"mutable run_optimization = false;","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1098,"value":" jobs = [\"5\",\"10\",\"25\"]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1111,"value":"vehicles = [\"5\",\"10\",\"20\",\"50\",\"100\"]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1163,"value":"useCases = [\n {name: \"Delivery\", value: \"delivery\"},\n {name: \"Service Worker\", value: \"service\"}\n ]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1175,"value":"objectives = [\n {name: \"Minimize Time\", value: \"duration\"},\n {name: \"Minimize Distance\", value: \"distance\"}\n]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2241,"value":"officeLocation = {\n let officeLL = panel_random.city.centroid.split(\",\");\n return (officeLL[0] + \",\" + officeLL[1]);\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2473,"value":"driverDepartureTime = {\n let departureEpoch = parseInt(selectedRoute2[0].steps[0].arrival) - parseInt(furthestEmployee.duration);\n let humanReadableTime = (new Date( departureEpoch *1000)).toLocaleString();\n return humanReadableTime;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1523,"value":"$36","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":924,"value":"$37","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2351,"value":"$38","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1279,"value":"routesFiltered = Object({\n type: 'FeatureCollection',\n features: routesGeoJson.features.filter(f => f.properties.vin === routeSelector.properties.vin)\n});","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2356,"value":"routesFiltered2 = Object({\n type: 'FeatureCollection',\n features: routesGeoJson2.features.filter(f => f.properties.vin === routeSelector.properties.vin)\n});","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2511,"value":"import { toc } from \"@nebrius/indented-toc\" // This component generates the Table of Contents","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W39","stdlib":"1","titling":"auto"}} 32:"$31:body"

MVRP Request ID: ${mvrpID?mvrpID.id:mvrpID}

Selected MVRP Request ID: ${mvrpID2?mvrpID2.id:mvrpID2}

For the driver to arrive at the first pickup location (the furthest employee) at time indicated in the table above, he will need to start from the office at ${driverDepartureTime}