2f:Tb38, ### Goal The goal of the POC is to demonstrate on a small data sample of student locations and buses that we can automatically assign students to buses and determine the optimal pickup sequences of student to meet the following criteria: * _"The students should spend minimal time in the bus to go to school."_ * _"The students should not wait too long for the bus to arrive at my pickup location."_ * _"The students should not reach school too much in advance before school actually starts."_ In this particular implementation students satisfaction are the main criteria when evaluating the quality of the transportation service. Note this notebook runs real code against NextBillion.ai' API. The purpose is to demonstrate one possible approach to solve the students transportation use-case given a set of constraints.
### Problem Statement Each School is assigned a zone. Students belonging to the zone will be picked up every morning and taken to school. The students are spread across the city or suburb within the defined zone. For this POC, 40 buses (minivans), starting from the school, are available to complete the students transportation. The drivers shift are from 06:00 to 09:00am. Each bus has a capacity of 6 seats (capacity configurable), driver excluded. The key criteria of a successful service is student satisfaction, as such the following needs to be taken into account when planning: * An accurate pickup time should be provide to each student to minimize the wait time of the bus. * Each student should be picked-up onward (starting with the furthest student) to school to minimize time spend in transport. * The buses should reach the school between 08:45 and 09:00am, just before the school starts, to minimize wait time at school.
### Approach The approach we will follow to propose one solution to this problem is described below: * Taking into account all students locations, available buses, and constraints (capacity, driver shift), compute the optimal route for each driver to pickup up to students. This will lead to groups of students assigned for pickup to different buses. * Compute for each group of students a Distance Matrix from school location to each students location to find the furthest student location. * For each bus, set the driver start location to the furthest student location and end location to the school. * Add an arrival time constraint at school from 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) 30:T6df,tempMVRPProblem = { if (run_optimization && inputData_a) { let location = ""; let loc_index = 0; let shipments = []; let jobs = []; let vehicles = []; let jobs_index = 0; customVRPInputs.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 <= customVRPInputs.pointsArray.length ) { location += "|"; } location += `${p.pickup_latitude},${p.pickup_longitude}`; jobs_index++; }); let timeOffset = new Date().getTimezoneOffset() * 60000; console.log("offset: ", timeOffset); let vehicles_index = 0; customVRPInputs.vehicleArray.forEach((f) => { if (vehicles_index < customVRPInputs.vehicleArray.length) { location += "|"; } location += `${f.latitude},${f.longitude}`; let start_time = inputData_a.shiftStart.toISOString().slice(0, -5) + "+03:00"; let end_time = inputData_a.shiftEnd.toISOString().slice(0, -5) + "+03:00"; let vehicle = { id: f.id, description: f.vin, start_index: loc_index, end_index: loc_index++, time_window: [ Date.parse(start_time) / 1000, Date.parse(end_time) / 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 }; } }31:T8de,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.description, 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 += "|" + furthestStudent.location[0] + "," + furthestStudent.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 }; } }32:T679,inputs = { let tomorrowStart = new Date(); tomorrowStart.setDate(new Date().getDate() + 1); tomorrowStart.setHours(6); tomorrowStart.setMinutes(0); 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); let startISOTime = new Date(tomorrowStart).toISOString().slice(0, -1) + "-04:00"; let endISOTime = new Date(tomorrowEnd).toISOString().slice(0, -1) + "-04:00"; 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: startISOTime }), shiftEnd: Inputs.datetime({ label: "Shift End", value: endISOTime }), 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: "6" }) }; return inputs; }33: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; }34: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; }35:[{"id":"1aa65d41117e06b5d7be10099a1380de1edbdd6df8ebb06bc2dc93aa8bff091e12fc923d0690e47531cedad8f65907025444a65d435da17419a422114da05339","href":"https://static.observableusercontent.com/files/1aa65d41117e06b5d7be10099a1380de1edbdd6df8ebb06bc2dc93aa8bff091e12fc923d0690e47531cedad8f65907025444a65d435da17419a422114da05339","name":"dubai_students@1.csv","type":"text/csv","lastModified":1723104263563.748,"size":5281},{"id":"c395e76332b6014b8c7e9b76880b625fe7705916b775c3c0046fb691002a9523e60fd922b95a3d8964fcb4707611a8dde37b8f2fc81dd181d67654a32ed74869","href":"https://static.observableusercontent.com/files/c395e76332b6014b8c7e9b76880b625fe7705916b775c3c0046fb691002a9523e60fd922b95a3d8964fcb4707611a8dde37b8f2fc81dd181d67654a32ed74869","name":"dubai_vehicles.csv","type":"text/csv","lastModified":1723104263563.748,"size":4418}] 2d:{"title":"Students Assignment & Transportation","cells":[{"id":0,"value":"![text](https://nextbillion.ai/wp-content/uploads/2022/01/logo-nb-colored.png)\n\n# Students Assignment & Transportation\n# Proof Of Concept\n","pinned":false,"hidden":false,"mode":"md"},{"id":2518,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":2517,"value":"toc(\"h2, h3, h4\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":2516,"value":"---\n","pinned":false,"hidden":false,"mode":"md"},{"id":508,"value":"$2f","pinned":false,"hidden":false,"mode":"md"},{"id":2882,"value":"viewof apiKey = secret2(\"apiKey\", {\n title: \"NextBillion.ai API Key\",\n submit: \"Set\",\n description:\n \"Contact support@nextbillion.ai to get your API key\"\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":525,"value":"### Loading Custom Data\nBelow we are loading fake randomly generated students and vehicles CSVs.\n

\n","pinned":false,"hidden":false,"mode":"md"},{"id":2024,"value":"viewof panel_random = Inputs.form(inputs, {\n template: (v) => htl.html`\n \n`\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":2783,"value":"dubai_students1 = await FileAttachment(\"dubai_students@1.csv\").csv({typed: \"auto\"})","pinned":true,"hidden":false,"mode":"ojs"},{"id":2785,"value":"dubai_vehicles = await FileAttachment(\"dubai_vehicles.csv\").csv({typed: \"auto\"})","pinned":true,"hidden":false,"mode":"ojs"},{"id":1236,"value":"---\n","pinned":false,"hidden":false,"mode":"md"},{"id":2559,"value":"customVRPInputs = {\n let retval = null;\n let pointsArray = [];\n let vehiclesArray = [];\n let employee = {};\n let vehicle = {};\n let employee_index = 1;\n let vehicle_index = 1;\n\n dubai_students1.forEach((row) => {\n employee = {\n id: employee_index++,\n pickup_longitude: row.lon,\n pickup_latitude: row.lat,\n business: `${row.student_id}`\n };\n pointsArray.push(employee);\n });\n dubai_vehicles.forEach((row) => {\n let location = row.start_location.split(\",\");\n vehicle = {\n id: vehicle_index++,\n vin: `${row.vin}`,\n latitude: location[0],\n longitude: location[1]\n };\n vehiclesArray.push(vehicle);\n });\n\n return { vehicleArray: vehiclesArray, pointsArray: pointsArray };\n\n console.log(vehiclesArray);\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":2490,"value":"### Problem Constraints\n\nHere we are setting a few constrains including the bus drivers shift, vehicle maximum capacity and objective function to use.\n","pinned":false,"hidden":false,"mode":"md"},{"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"},{"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"},{"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 students 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"},{"id":190,"value":"$30","pinned":false,"hidden":false,"mode":"ojs"},{"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"},{"id":221,"value":"mvrpID = {\n const mvrpRequest = {\n description: `MVRP Tutorial ${new Date().toISOString()}`,\n locations: tempMVRPProblem.locations,\n vehicles: tempMVRPProblem.vehicles,\n jobs: tempMVRPProblem.jobs,\n options: {\n objective: {\n travel_cost: inputs.objectiveFunction.value.value,\n minimise_num_depots: false\n }\n }\n };\n console.log(JSON.stringify(mvrpRequest));\n let resp = axios\n .post(`https://api.nextbillion.io/optimise-mvrp?key=${apiKey}`, mvrpRequest)\n .then((response) => response.data)\n .catch((response) => response.data);\n return resp;\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":348,"value":"viewof jobID = htl.html`MVRP Request ID: ${\n mvrpID ? mvrpID.id : mvrpID\n}`","pinned":false,"hidden":false,"mode":"ojs"},{"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

","pinned":false,"hidden":false,"mode":"md"},{"id":720,"value":"html`IMPORTANT
If you get a status message indicating Job still processing just wait a few seconds and press this button.

${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs"},{"id":242,"value":"checkStatus = {\n mvrpID;\n await visibility();\n requestResult;\n yield Promises.delay(5000).then(() => {\n let mvrpResult = axios({\n method: \"GET\",\n url: `https://api.nextbillion.io/optimise-mvrp/result?id=${mvrpID.id}&key=${apiKey}`,\n headers: {}\n })\n .then((res) => res.data)\n .catch((error) => {\n let mvrpResult = {\n status: \"Error\",\n error: JSON.stringify(error)\n };\n });\n return mvrpResult;\n });\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":631,"value":"kpiStats = {\n let deliveryCount = 0;\n checkStatus.result.routes.forEach((rte, rteidx) => {\n rte.steps.forEach((step, index) => {\n if (step.id && step.id != 99) {\n ++deliveryCount;\n }\n });\n });\n return {\n totalDistance: `${(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"},{"id":2414,"value":"### KPIs for all buses","pinned":false,"hidden":false,"mode":"md"},{"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"},{"id":605,"value":"---\n### Render the intermediate Solution on a Map\nThis maps shows groups of students assigned to each bus. This is not the final students pickup sequence. This is an intermediary step.\n","pinned":false,"hidden":false,"mode":"md"},{"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"},{"id":1676,"value":"green marker is the vehicle start location (the school) and red markers are students pickup locations. You can click on the markers to get details.","pinned":false,"hidden":false,"mode":"md"},{"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"},{"id":1862,"value":"Inputs.table(selectedRoute[0].steps, {\n columns: [\"type\", \"arrival\", \"load\", \"description\", \"duration\"],\n header: {\n type: \"Task\",\n arrival: \"Arrival Time\",\n load: \"Nbr Loaded\",\n description: \"Student\",\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) =>\n new Date(x * 1000).toLocaleTimeString([], {\n hour: \"2-digit\",\n minute: \"2-digit\",\n hour12: false\n }),\n duration: (x) => (x / 60).toFixed(1) + \" mn\"\n },\n rows: 10,\n maxWidth: 600,\n multiple: false,\n required: false\n})\n\n//.toLocaleString(\"en-US\", { timeZone: \"Asia/Calcutta\" });","pinned":false,"hidden":false,"mode":"ojs"},{"id":1898,"value":"selectedRoute = checkStatus.result.routes.filter(f => f.description === routeSelector.properties.vin)","pinned":false,"hidden":false,"mode":"ojs"},{"id":2359,"value":"selectedRoute2 = checkStatus2.result.routes.filter(\n (f) => f.description === routeSelector2.properties.vin\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":2743,"value":"### Visualize All Routes Using Optimizer \nIf you want to visualize the optimized routing solution along with the sequencing of the stops, the NB.ai Optimization Tester tool is available at NB.AI MVRP Playground App","pinned":false,"hidden":false,"mode":"md"},{"id":2164,"value":"---\n### Optimize each driver journey to meet students requirements\n\nHere we want to set the furthest student from the school to be the start location of the driver journey. This will ensure the students are picked in the appropriate order from the furthest to the closest, thus to minimize the time each student spent in the bus. ","pinned":false,"hidden":false,"mode":"md"},{"id":2173,"value":"### Distance Matrix Computation\n\nFor each optimized group of students created by the optimizer (all the students to be picked up by a given driver), we will assess who is the furthest students. As such we compute a Distance Matrix where the origin is the school and the destinations are the student locations.","pinned":false,"hidden":false,"mode":"md"},{"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 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 //console.log(destinations);\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":2214,"value":"### Furthest Student\n\nHere we apply some basic logic by parsing the distance matrix output to pick the employee for which the distance is the furthest student from the school.","pinned":false,"hidden":false,"mode":"md"},{"id":2218,"value":"furthestStudent = {\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"},{"id":2418,"value":"---\n### Build the Problem Inputs for each Student group\nNow we build the needed JSON objects that will be submitted to the MVRP endpoint for each student group. We are as before building the `jobs` which each consist of students pickup locations in the city region with the main difference that we apply the optimizer to a subset i.e one single student group. In addition we set new constraints:\n* the driver start location is the furthest student location to ensure the furthest student is picked up first.\n* an artificial job (with no pickup) is created at the school location with a pickup time window constraint set from 08:45 to 09:00 am.","pinned":false,"hidden":false,"mode":"md"},{"id":2287,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"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"},{"id":2292,"value":"viewof jobID2 = htl.html`
Selected MVRP Request ID: ${mvrpID2?mvrpID2.id:mvrpID2}
`","pinned":false,"hidden":false,"mode":"ojs"},{"id":2313,"value":"html`
If you get a status message indicating Job still processing just wait a few seconds and press this button.

${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs"},{"id":2341,"value":"checkStatus2 = {\n mvrpID2;\n //await visibility();\n requestResult;\n yield Promises.delay(5000).then(() => {\n let mvrpResult2 = axios({\n method: \"GET\",\n url: `https://api.nextbillion.io/optimise-mvrp/result?id=${mvrpID2.id}&key=${apiKey}`,\n headers: {\n } \n }).then(res=>res.data)\n .catch(error => {\n let mvrpResult = {\n status: 'Error', \n error: JSON.stringify(error)\n }\n });\n return mvrpResult2;\n })\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":2344,"value":"kpiStats2 = {\n let deliveryCount = 0;\n checkStatus2.result.routes.forEach((rte, rteidx) => {\n rte.steps.forEach((step, index) => {\n if (step.id && step.id != 99) {\n ++deliveryCount;\n }\n });\n });\n return {\n totalDistance: `${(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"},{"id":2346,"value":"---\n### Render Improved Solution","pinned":false,"hidden":false,"mode":"md"},{"id":2348,"value":"viewof routeSelector2 = Inputs.select(\n routesGeoJson2.features.filter((f) => f.geometry.type === \"LineString\"),\n {\n multiple: false,\n label: \"Selected route to view\",\n format: (x) => x.properties.vin,\n value: (t) => t.properties.vin\n }\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":2354,"value":"{\n return geoviewer(routesFiltered2, {width:\"100%\",height:\"550\",col:'#8058a0',fillOpacity:.9,lineWidth:3,style:'nbai',apiKey:apiKey});\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":2361,"value":"Inputs.table(selectedRoute2[0].steps, {\n columns: [\"type\", \"arrival\", \"load\", \"description\", \"duration\"],\n header: {\n type: \"Task\",\n arrival: \"Arrival Time\",\n load: \"Nbr Loaded\",\n description: \"Student\",\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) =>\n new Date(x * 1000).toLocaleTimeString([], {\n hour: \"2-digit\",\n minute: \"2-digit\",\n hour12: false\n }),\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"},{"id":2467,"value":"viewof driverDeparture = htl.html`For the driver to arrive at the first pickup location (the furthest pickup location) at time indicated in the table above, he will need to start from the school at ${driverDepartureTime}`","pinned":false,"hidden":false,"mode":"ojs"},{"id":2746,"value":"### Visualize Single Route Using Optimizer \nIf you want to visualize the optimized routing solution along with the sequencing of the stops, the NB.ai Optimization Tester tool is available. \n\nNB.AI MVRP Playground App (should automatically load the current solution)","pinned":false,"hidden":false,"mode":"md"},{"id":2443,"value":"### Conclusion\n\nUsing a multi-steps approach where the optimizer is first used as a pre-processing step to create students groups, running the distance matrix inside each group to assess who is the furthest student and finally running the optimization again inside each group, we manage to achieve the set requirements as highlighted in the table above, which ultimately should lead to student satisfaction.\n\nBy following this multi-step approach we were able to:\n* Provide a time to each student for the pickup.\n* Minimize the time spend in the bus.\n* Arrive a few minutes before the actual school starts.","pinned":false,"hidden":false,"mode":"md"},{"id":1000,"value":"\n\n\n\n---\n### Additional Tools \nIf you want to visualize the optimized routing solution along with the sequencing of the stops, the NB.ai Optimization Tester tool is available. \n\nNB.AI MVRP Playground App (should automatically load the current solution)\n- Select the `SOLUTION` tab\n- Enter the same NB.ai API key used for the MVRP solution `${apiKey}`\n- Paste the MVRP Request ID into the input box `${mvrpID.id}`\n- Press the `Load` button","pinned":false,"hidden":false,"mode":"md"},{"id":341,"value":"### Other Dependencies","pinned":false,"hidden":false,"mode":"md"},{"id":1139,"value":"import {nextbillion,cities,secret2} from \"@yvannbai/nextbillion-ai-helpers\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":334,"value":"axios = {\n let axios = await require('axios');\n return axios;\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":352,"value":"styling = html`\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":849,"value":"geoviewer = require(\"https://d12qcqjlhp2ahm.cloudfront.net/nb_geoverview2.js\").then((f) => f.view)","pinned":false,"hidden":false,"mode":"ojs"},{"id":1613,"value":"html`${viewof requestResult}`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1005,"value":"viewof requestResult = Inputs.button(\"Request Result\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1686,"value":"mutable inputData_a = null","pinned":false,"hidden":false,"mode":"ojs"},{"id":2051,"value":"mutable run_optimization = false;","pinned":false,"hidden":false,"mode":"ojs"},{"id":1098,"value":" jobs = [\"5\",\"10\",\"25\"]","pinned":false,"hidden":false,"mode":"ojs"},{"id":1111,"value":"vehicles = [\"5\",\"10\",\"20\",\"50\",\"100\"]","pinned":false,"hidden":false,"mode":"ojs"},{"id":1163,"value":"useCases = [\n {name: \"Delivery\", value: \"delivery\"},\n {name: \"Service Worker\", value: \"service\"}\n ]","pinned":false,"hidden":false,"mode":"ojs"},{"id":1175,"value":"objectives = [\n {name: \"Minimize Time\", value: \"duration\"},\n {name: \"Minimize Distance\", value: \"distance\"}\n]","pinned":false,"hidden":false,"mode":"ojs"},{"id":2241,"value":"officeLocation = {\n return \"25.178899,55.306185\";\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":2473,"value":"driverDepartureTime = {\n let departureEpoch =\n parseInt(selectedRoute2[0].steps[0].arrival) -\n parseInt(furthestStudent.duration);\n let humanReadableTime = new Date(departureEpoch * 1000).toLocaleString();\n return humanReadableTime;\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":1523,"value":"$32","pinned":false,"hidden":false,"mode":"ojs"},{"id":924,"value":"$33","pinned":false,"hidden":false,"mode":"ojs"},{"id":2351,"value":"$34","pinned":false,"hidden":false,"mode":"ojs"},{"id":1279,"value":"routesFiltered = Object({\n type: 'FeatureCollection',\n features: routesGeoJson.features.filter(f => f.properties.vin === routeSelector.properties.vin)\n});","pinned":false,"hidden":false,"mode":"ojs"},{"id":2356,"value":"routesFiltered2 = Object({\n type: 'FeatureCollection',\n features: routesGeoJson2.features.filter(f => f.properties.vin === routeSelector.properties.vin)\n});","pinned":false,"hidden":false,"mode":"ojs"},{"id":2511,"value":"import { toc } from \"@nebrius/indented-toc\" // This component generates the Table of Contents","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W35","stdlib":"1","titling":"auto"}

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