Table of Contents
- When to use Review
- Dispatch and Live Dispatching at a glance
- How Review gets populated
- Page layout
- The top bar
- Filtering the plan
- Drivers panel
- Trips panel
- Footer totals
- Recommendation details
- Alternate recommendations
- Locking and rejecting in Live Dispatching
- Diagnostics
- Downloading the plan
- Tips and best practices
When to use Review
- After any bulk optimization run, to validate the plan before saving to your TMS
- To open Live Dispatching and work from the latest shared plan without running anything yourself
- To compare alternate driver options for a trip that matters
- To evaluate the plan’s profit, mileage, and service performance
- To understand why certain trips or drivers went uncovered
- To re-run optimization with different settings and compare the results
Dispatch and Live Dispatching at a glance
Use the dropdown at the top left of Review to switch between the two modes. They look the same but work differently.
| Dispatch (bulk optimization) | Live Dispatching | |
|---|---|---|
| What it is | The result of a bulk optimization you launched. | A shared plan that LoadAi refreshes automatically in the background. |
| How it’s created | You run a bulk optimization, choosing the Engine Mode, drivers, and trips. | An administrator saves a configuration (Engine Mode, drivers, trips). LoadAi then refreshes the plan over time. |
| Who sees it | The plan you created. | Everyone in your organization sees the same plan. |
| How it stays current | It doesn’t. It’s a snapshot of the moment you ran it. Re-Optymize replaces it with a new run. Stale data warnings will appear as trips update. | Each run starts from the previous plan, reconciled with your latest TMS data, and improves it. The time of the latest run is shown. |
| Re-running | Re-Optymize runs the engine again with updated settings. | Runs automatically about every 10-15 minutes. Authorized users can also trigger a run. |
| Lock and reject | Not currently supported. | Lock a pairing to keep it, or reject one with a reason. Future runs honor both. |
| Rejection history | Not currently supported. | Rejections opens a history of rejected recommendations. |
| Filter pills | Selected, Recommended, Not Matched, Repositions, Relay Created. Brokered when applicable. | The same, plus Locked. |
| Save to TMS | Commits the selected assignments. | Commits the selected assignments. Locking and rejecting don’t change who can save. |
| Best for | A deliberate planning run where you control the inputs and compare results. | Opening the page and always having a current best plan. |
For how Live Dispatching is set up and how locks and rejections work, see Live Optimization.
How Review gets populated
Dispatch. Review is populated when a sandbox plan exists. A plan is created by running a bulk optimization.
- After a bulk optimization finishes, Review can be opened automatically from the notification.
- You can also reach it any time from Planning > Review in the navigation panel.
- If no plan exists, Review will prompt you to start a bulk optimization.
Only one sandbox plan exists at a time. To start fresh, delete the current plan or run a new optimization.
For how to launch a run, see Bulk Optimization.
Live Dispatching. Choose Live Dispatching in the dropdown. It’s available only when Live Optimization is enabled for your account. If you don’t see it, contact your Customer Success Manager.
Page layout
Review uses the same layout as the Dispatch page, because a Review plan is a preview of what Dispatch will look like once you save it. From top to bottom:
| Area | What it does |
|---|---|
| Top bar | Mode selection, plan actions, diagnostics, and the view toggle |
| Filter bar | AI Filter, the filter pills, and Add Filter |
| Drivers panel | Each driver’s schedule as a timeline, with recommended trips added |
| Trips panel | One row per recommended trip |
| Footer | The number of drivers and trips shown, with the plan’s financial totals |
The Drivers and Trips panels both follow your filters and selection. You can expand either panel to see more of it.
The top bar
Mode. The dropdown at the top left selects Dispatch or Live Dispatching. Depending on your subscription, Dispatch may also include Partial Load Consolidation and LTL Load Consolidation plans.
More Actions. Three vertical dots represent the overflow actions. Here you can Download Report to get a file of the current dispatch plan or Delete Plan which clears the sandbox.
Info Button. The information button includes the diagnostics of the bulk optimization, the Rune time, and the Optimization Mode.
View toggle. Three icons at the top right switch the layout between Drivers only, Drivers and Trips, and Trips only.
Actions.
| Action | Where | What it does |
|---|---|---|
| Re-Optymize | Dispatch | Runs the engine again over the same work with updated settings, replacing the current plan. |
| Optimize (refresh icon) | Live Dispatching | Opens the Live Dispatching configuration and runs a new optimization. |
| Rejections | Live Dispatching | Opens the history of rejected recommendations. |
| Save to TMS | Both | Commits the selected assignments. They become real assignments, go to your TMS, and leave Review. It’s disabled until you select something. |
| ⋮ menu | Both | Delete plan and Download Report |
| ⓘ (info) | Both | View diagnostics, Run Time, and the Engine Mode used. |
Save to TMS is the commit point. Review the plan fully before using it.
Checking the Engine Mode is worth doing before you judge a plan. The mode determines what the engine was allowed to do: whether relays were permitted, how far ahead it looked, and so on. A plan that looks disappointing is often a plan that was run under a restrictive mode.
Filtering the plan
The filter bar sits under the top bar.
Pills. Each pill is a count and a filter at the same time. Select one or more pills to filter both panels, and the numbers update. Every count has two numbers: drivers and trips.
| Pill | Shows |
|---|---|
| Selected | The drivers and trips you’ve selected. |
| Locked | Locked pairings. Live Dispatching only. |
| Recommended | Current recommendations. This is selected by default. |
| Not Matched | Drivers and trips the engine couldn’t pair, as two separate counts. |
| Repositions | Trips that move a driver without freight to rebalance. |
| Relay Created | Trips where the engine recommended a driver handoff. |
| Brokered | Trips recommended for a third-party carrier. Hidden when nothing is brokered. |
For example, “Recommended: 54 drivers – 118 trips” means 54 drivers were planned on 118 recommended trips.
AI Filter. Describe the filter you want in plain English and it will attempt to match the filters.
Add Filter. Add more filters to narrow the plan further. Filters apply to both panels. If you need an additional filter, let your Customer Success Manager know.
| Filter | Use |
|---|---|
| Has Recommendation | Applied as Yes by default. Switch it off to look at work that went uncovered. |
| Is Relay | Show only recommendations involving a driver handoff |
| Rating | Filter by recommendation quality score |
| Tour Miles | Filter by total miles of the recommended tour |
| Fleet | Filter by driver fleet |
| Operation Type | Filter by driver operation type |
| Solo-Team | Solo or team drivers |
Select Reset Filters to clear them. ⚠ Confirm whether Has Recommendation and the rest of the old list still apply now that Recommended is a pill.
Drivers panel
The Drivers panel is a timeline: one row per driver across a multi-day view, with a vertical line marking the current time. It works like the Gantt on the Dispatch page.
Each row shows the driver’s status, name, and location, with tractor and trailer links, followed by their work laid out in time order. A Select All button at the top selects every driver.
- Recommended trips appear on the timeline next to the work the driver already has, so you can see what the plan adds.
- Trips a driver already has, and any time off, appear in order with the recommendations. This includes drivers with several gaps, where a recommendation can place trips in more than one.
- Trips planned in LoadAi are highlighted so they’re easy to spot. ⚠ Confirm the highlight.
- Badges flag issues: delay badges show projected lateness with the time, and warning badges such as Low Utilization flag planning issues.
- Zoom with − and +, return to the default with Reset, or jump to the current time with Now.
- Use the ⚙ settings icon to choose what appears on the timeline, the same options as on the Dispatch page.
Driver details
Select a driver to open their panel. It shows:
- Driver summary: name, employment and operation type, current location and duty status
- Weekly targets: miles, revenue, and earnings progress against target, with a day-by-day breakdown
- Schedule: the work the driver already has and the recommended trips. If their current schedule changes you can toggle the Initial Schedule to see the difference.
- Recommendations: the trips this plan proposes adding, with times, locations, and mileage
Trips panel
One row per recommended trip. Use Search to find a trip, and the ⚙ icon to choose which columns appear.
Where the engine has built a tour, several trips run back-to-back by one driver, each trip still appears as its own row. Selecting any one of them selects the whole tour. Tours are accepted or rejected together.
Columns
| Column | Description |
|---|---|
| Trip ID | The recommended trip. Click to open the detail panel. |
| Assign to | The driver the engine recommends |
| Status | Where the recommendation is in its lifecycle (see below) |
| Rating | Recommendation quality, scored out of 5 |
| Tractor / Trailer | Recommended equipment, where that data is available |
| Customer | Customer on the trip |
| Revenue Code | Your pricing agreement or business unit reference for the trip |
| Has Relay | Whether a driver handoff is involved |
| Deadhead % | Empty miles as a share of total miles |
| Loaded Miles | Revenue-generating miles |
| Weight | The weight of the load |
| From City / To City | Trip origin and destination |
| Profit | Modeled profit. This can be negative. |
| Revenue per mile | Revenue rate for the trip |
| Trip Start Zone / Trip End Zone | Optimization zones for each end of the trip |
| Start Time / End Time | Planned start and finish, in local time |
In Live Dispatching, each row also has Lock and Reject buttons, next to the assign button.
Recommendation status
A recommendation moves through states as you work:
| Status | Meaning |
|---|---|
| Recommended | Proposed by the engine, not yet acted on. Every row starts here. |
| Pending Assignment | You’ve accepted it. It’s queued to become an assignment. |
| Assigned | The pairing is an assignment. |
When you save the plan, assignments are pushed to your TMS and leave Review.
Footer totals
The footer summarizes what you’re looking at:
| Figure | What it means |
|---|---|
| Drivers and trips shown | The number of drivers and trips under your current filters. |
| Revenue | Total revenue. The figure in brackets is revenue per total mile, loaded and empty miles combined. |
| Profit | Revenue less modeled cost, with average profit per driver per day in brackets. |
| Loaded Miles | Revenue-generating miles. |
| Deadhead | Empty miles, with deadhead shown as a percentage of total miles. |
The totals recalculate for your filters and selection, so you can evaluate one fleet, one region, or one driver’s worth of work in isolation.
Note on revenue per mile: the bracketed figure divides revenue by total miles, not loaded miles. If you’re comparing it to a loaded-mile rate from elsewhere, the numbers will differ. This is expected.
Recommendation details
Click any trip to open the detail panel. This is where you evaluate a recommendation properly.
Header
The top of the panel shows the trip’s profit, its appointment windows and estimated arrival and departure times, origin and destination, total distance and duration, the load and customer, and a priority badge where one applies.
The recommended tour
Below the header you’ll see the full tour the recommendation implies, every leg the driver would run, not just the trip you clicked. Each leg shows its trip and load, the customer, departure and arrival times, locations, distance, and duration.
This matters: accepting a recommendation accepts the whole tour.
Tour summary metrics
Five figures summarize the tour:
| Metric | What it measures |
|---|---|
| Deadhead miles | Empty miles across the tour |
| Profit | Modeled profit for the tour |
| Utilization | This is a measure of both drive and duty time as well as weekly targets |
| Delay | Projected lateness against appointments |
| Rules | How many of the five matching requirements this pairing satisfies |
The five matching requirements
The requirements met figure counts these:
- Equipment compatibility: the trailer type matches what the load requires
- Driver area preferences: the work falls in areas the driver prefers, and avoids those they don’t
- Certifications: the driver holds the endorsements the load needs
- Shift preference: the work fits the driver’s day or night preference
- Driver load matching rules: any matching rules you’ve configured
A pairing showing 4 of 5 is workable. It simply means one preference or rule wasn’t satisfied. Open the detail to see which.
How the rating works
The Rating is scored out of 5. Each of the five tour metrics above contributes one point, scaled by how good that metric is for the pairing.
The score is deliberately balanced, so no single factor dominates. A recommendation with excellent profit but poor appointment adherence won’t outrank a well-rounded one on rating alone.
If your priority this week is specific (covering deadhead, catching drivers up on miles, protecting service), don’t rely on rating alone. Sort the alternates against that priority instead.
Alternate recommendations
Expand Alternate Recommendations to see other drivers who could take the trip. Each alternate shows the candidate, their rating, and the full tour they’d run.
Optimization mode
The mode selector at the top of the section controls the algorithm used to generate alternates. Changing it re-runs One to Many Optimization.
Sorting alternates
Alternates are ranked by overall rating by default. Use the sort control to rank them against a single objective instead:
| Sort by | Ranks by | Check it against |
|---|---|---|
| High Rating | Overall balanced score | The star rating |
| High Profit | Most profitable tour | The profit figure |
| Low Deadhead Miles | Fewest empty miles | The deadhead figure |
| Low Delay Time | Best appointment adherence | The delay figure |
| High Requirements Met | Most matching requirements satisfied | The requirements figure |
| Early Assignment | Soonest the driver can start | The ETAs on each leg |
| High Tour Miles | Most total miles for the driver | The leg distances added together |
Sorting doesn’t change any recommendation’s rating. It only changes the order you see them in.
A High MIP Term option also appears in this menu. It’s a troubleshooting aid used when investigating how the engine ranked candidates, and isn’t needed for normal planning.
Changing one recommendation is not a local change. Because the engine builds tours and balances work across the whole population, moving one trip to a different driver can unwind several other recommendations. The notices tell you exactly what would be affected before you commit to it.
Locking and rejecting in Live Dispatching
In Live Dispatching, you can tell the engine your decisions, and future runs respect them:
- Lock a driver–trip pairing to keep it through every refresh. Only that pairing is locked, not the rest of its tour.
- Reject a pairing you don’t want, with a required reason. The engine won’t recommend that trip to that driver again, though it can still go to other drivers.
- Locks and rejections are shared. Everyone using Live Dispatching sees them.
- Rejections in the top bar opens the history of past rejections, with a total count.
- To lock or reject several rows at once, select them and choose Lock or Reject.
Locking and rejecting never block Save to TMS. See Live Optimization for full details, including who can lock and reject.
Diagnostics
Diagnostics explains why trips or drivers went uncovered. Each reason carries a percentage showing how often it blocked an assignment across the run.
Typical reasons include deadhead beyond the allowed threshold, a pickup time that falls outside the driver’s available window, driver unavailability, and hours-of-service limits.
A large Not Matched count usually points to a constraint or a data problem rather than a shortcoming in the plan. Diagnostics is the fastest way to find out which.
Tips and best practices
- Check the Engine Mode first. It tells you what the engine was allowed to do, which often explains the plan you’re looking at.
- Use Network Balance beforehand to decide which zones need attention before you run.
- Work the Not Matched count early. Diagnostics usually points to a constraint or a data problem you can fix and re-run, rather than something you need to plan around.
- Use the pills. Select Recommended and Relay Created together, for instance, to review only recommendations involving a handoff.
- Select assignments to scope the totals. The footer recalculates for whatever you’ve selected, so you can evaluate one fleet, one region, or one driver’s worth of work in isolation.
- Compare runs. In Dispatch mode, try more than one Engine Mode and compare the results rather than accepting the first plan.
- Use Live Dispatching for the current best plan. Lock the pairings you like and reject the ones that don’t work so the next refresh builds on your decisions.
- Save to TMS last. It’s the commit point, and assignments leave Review at that moment.