Skip to content

Research ยท October 2026 edition

On-Site Robot Training Jobs: Labs, Depots and Test Tracks

Of the 1,150 robot training roles open on 31 Oct 2026, 55.1% called for on-site or hybrid work, 6.9% were remote and 38.0% gave no arrangement; autonomous vehicle and truck operation was the biggest on-site group, and most on-site roles were in the United States, Japan and Germany.

Read the month's story: Robot Training Jobs in October 2026: The Month in Data

Not final yet. The month has ended, yet stragglers in the data may still land up to 3 Nov 2026; after that the edition is locked permanently.

Period
1 Oct 2026 to 31 Oct 2026
31 days, UTC
Calculated
1 Nov 2026, 06:05 UTC
Settles when locked
Where it comes from
Daily reads of the public job boards run by the robot training employers and programs this site follows, using each open role's title, location text and the workplace field its board provides.
Method
Revision 1. Terms defined
CC BY 4.0, citing this report

Each figure here comes as CSV and JSON files, emailed to your work address.

What the data shows

  1. 1

    On 31 Oct 2026, 634 of the 1,150 open robot training roles (55.1%) had to be worked on-site or hybrid. 6.9% were remote, and 38.0% gave no arrangement.

    Figure 1
  2. 2

    The title, location and workplace field are all silent on arrangement for 437 open roles, and 38.9% of those sit at Tesla. The remote share should be read as a floor.

    Figure 2
  3. 3

    Among kinds of work with enough open roles to stand on their own, Robot Teleoperation leans furthest toward on-site work (74.4%, 61 roles). 7 kinds of work have at least one on-site role.

    Figure 1
  4. 4

    On-site roles come from 47 employers. TSMG posts the largest number: 503, or 79.3% of the on-site total.

    Figure 2
  5. 5

    Of the 51 countries with on-site openings, the United States accounts for the largest number (246, 38.8%); the single place named most often is the San Francisco Bay Area, at 39 roles.

    Figure 3
  6. 6

    Going by job titles, on-site roles cluster most in autonomous vehicle and truck operation (326 roles, 51.4%), followed by robot data collection (150, 23.7%).

    Figure 5
  7. 7

    On employers tracked the whole time, the on-site count was 568 on 1 Oct 2026 and 605 on 31 Oct 2026 (+6.5%); 69 new on-site roles were posted during the window.

    Figure 7

Figure 1

Where the work is done, kind by kind

Which kinds of robot training work need you on location instead of at home?

Kind of workOn-site or hybridRemoteNot stated
All kinds of work
55.1%
6.9%38.0%6341,150
Robot Teleoperation
74.4%
2.4%23.2%6182
Egocentric & Wearable Capture
73.9%
25.4%0.7%99134
AV & Drone Operators
63.1%
0.4%36.6%343544
Robot Task Evaluation & QA
44.4%
22.2%33.3%49
Data & Demonstration Collection
41.7%
26.0%32.3%53127
Robot Fleet Operations
29.5%
0.0%70.5%70237
Embodied-AI Data Annotation
23.5%
35.3%41.2%417

Base: 634 on-site or hybrid out of 1,150 roles still open at the close of 31 Oct 2026.

A role counts as on-site or hybrid when its title or location points to a place ('Onsite', 'On-site London', 'In-person', 'Hybrid'), or when a board with its own workplace field (Lever, Ashby) marks it as not remote. It counts as remote when the title or location says so, or when the board flags it remote. Titles like 'Remote Driver' or 'Remote Assistance Operator' describe how a robot or vehicle is steered, not where the person sits, so they are not taken as remote.

A role is not stated when its title, location and workplace field are all silent on where the work happens. Most robot training roles are tied to a lab, a depot or a test route, yet silence is never counted as on-site. That makes both the on-site and the remote shares floors, not guesses about the silent roles.

Hybrid and on-site share one column, since listings give no field that tells them apart.

A row with under 7.3 open roles carries a small-sample mark and is left unranked: under that count, the average across all rows says more than the row's own open roles do (a weight fitted on the counts themselves).

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 2

Where the work is done, employer by employer

Which employers post robot training roles you have to show up for in person?

EmployerOn-site or hybridRemoteNot stated
TSMG
100.0%
0.0%0.0%503503
Serve Robotics
100.0%
0.0%0.0%1616
Aurora Innovation
100.0%
0.0%0.0%99
Dyna Robotics
100.0%
0.0%0.0%99
Genesis AI
100.0%
0.0%0.0%66
ISEE
100.0%
0.0%0.0%66
Waabi
100.0%
0.0%0.0%66
Collaborative Robotics
100.0%
0.0%0.0%55
Saronic
100.0%
0.0%0.0%55
Scout AI
100.0%
0.0%0.0%55
Coco Robotics
100.0%
0.0%0.0%44
Monumental
100.0%
0.0%0.0%44
Shield AI
100.0%
0.0%0.0%44
Wayve
100.0%
0.0%0.0%44
AeroVect
100.0%
0.0%0.0%33
FS Studio
100.0%
0.0%0.0%33
1X Technologies
100.0%
0.0%0.0%22
Applied Intuition
100.0%
0.0%0.0%22
Cota Robotics
100.0%
0.0%0.0%22
Dexmate
100.0%
0.0%0.0%22
Gecko Robotics
100.0%
0.0%0.0%22
Havoc AI
100.0%
0.0%0.0%22
Humanoid
100.0%
0.0%0.0%22
PlusAI
100.0%
0.0%0.0%22
Rhoda AI
100.0%
0.0%0.0%22
Sereact
100.0%
0.0%0.0%22
Tutor Intelligence
100.0%
0.0%0.0%22
Nuro
14.3%
0.0%85.7%17
Arctic Engines
25.0%
25.0%50.0%14
Welo Data
33.3%
66.7%0.0%13
Sunday Robotics
50.0%
50.0%0.0%12
120ft
100.0%
0.0%0.0%11
Adexis
100.0%
0.0%0.0%11
Ambi Robotics
100.0%
0.0%0.0%11
Appen
100.0%
0.0%0.0%11
Droyd
100.0%
0.0%0.0%11
Enact Intelligence
100.0%
0.0%0.0%11
Generalist AI
100.0%
0.0%0.0%11
May Mobility
100.0%
0.0%0.0%11
Mind Robotics
100.0%
0.0%0.0%11
Nidus
100.0%
0.0%0.0%11
Physical Intelligence
100.0%
0.0%0.0%11
Pickle Robot
100.0%
0.0%0.0%11
Pyka
100.0%
0.0%0.0%11
Reliable Robotics
100.0%
0.0%0.0%11
Skyways
100.0%
0.0%0.0%11
Telexistence
100.0%
0.0%0.0%11
Tesla
0.0%
0.0%100.0%0170
Amazon Prime Air
0.0%
0.0%100.0%059
Alignerr
0.0%
100.0%0.0%041
HireArt
0.0%
0.0%100.0%025
Transdev
0.0%
0.0%100.0%025
Claru
0.0%
100.0%0.0%016
Avride
0.0%
0.0%100.0%015
Gatik
0.0%
0.0%100.0%011
Symbotic
0.0%
0.0%100.0%011
DoorDash Labs
0.0%
0.0%100.0%010
Figure AI
0.0%
0.0%100.0%010
Anduril Industries
0.0%
0.0%100.0%09
Moove
0.0%
0.0%100.0%09
Zipline
0.0%
0.0%100.0%08
Robot.com (Kiwibot)
0.0%
0.0%100.0%07
Kodiak AI
0.0%
0.0%100.0%06
Latitude AI
0.0%
16.7%83.3%06
micro1
0.0%
100.0%0.0%06
Waymo
0.0%
40.0%60.0%05
Skild AI
0.0%
0.0%100.0%04
Nimble Robotics
0.0%
0.0%100.0%03
Perle
0.0%
100.0%0.0%03
Skyports Drone Services
0.0%
0.0%100.0%03
Torc Robotics
0.0%
33.3%66.7%03
Beep
0.0%
0.0%100.0%02
Bot Auto
0.0%
0.0%100.0%02
Dexory
0.0%
0.0%100.0%02
Instawork
0.0%
0.0%100.0%02
Keywords Studios
0.0%
50.0%50.0%02
Motional
0.0%
0.0%100.0%02
Neros
0.0%
0.0%100.0%02
Performance Drone Works
0.0%
50.0%50.0%02
RoboForce
0.0%
0.0%100.0%02
Stack AV
0.0%
0.0%100.0%02
Aerospace engineering firm
0.0%
0.0%100.0%01
Allen Control Systems
0.0%
0.0%100.0%01
CHAOS Industries
0.0%
0.0%100.0%01
Flexdrive (Lyft)
0.0%
0.0%100.0%01
Foundation Robotics
0.0%
0.0%100.0%01
Innodata
0.0%
100.0%0.0%01
Manna
0.0%
0.0%100.0%01
Matternet
0.0%
0.0%100.0%01
Maven Robotics
0.0%
0.0%100.0%01
Mindy Support
0.0%
100.0%0.0%01
Pronto
0.0%
0.0%100.0%01
Rugged Robotics
0.0%
0.0%100.0%01
Saildrone
0.0%
0.0%100.0%01
Sama
0.0%
0.0%100.0%01
Starship Technologies
0.0%
0.0%100.0%01
TELUS Digital AI Community
0.0%
100.0%0.0%01
Vay
0.0%
0.0%100.0%01
Wing
0.0%
0.0%100.0%01
Workstream
0.0%
0.0%100.0%01

Base: 1,150 roles still open at the close of 31 Oct 2026, grouped by the employer posting them. These are listings, not hires.

A role counts as on-site or hybrid when its title or location points to a place ('Onsite', 'On-site London', 'In-person', 'Hybrid'), or when a board with its own workplace field (Lever, Ashby) marks it as not remote. It counts as remote when the title or location says so, or when the board flags it remote. Titles like 'Remote Driver' or 'Remote Assistance Operator' describe how a robot or vehicle is steered, not where the person sits, so they are not taken as remote.

A role is not stated when its title, location and workplace field are all silent on where the work happens. Most robot training roles are tied to a lab, a depot or a test route, yet silence is never counted as on-site. That makes both the on-site and the remote shares floors, not guesses about the silent roles.

None of the rows with open roles carries a small-sample mark: they vary so widely that each one stands on its own from the first.

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 3

On-site robot training roles, country by country

Which countries have in-person robot training openings?

CountryOn-site rolesShare of on-site roles
United States
246
38.8%San Francisco Bay Area, Atlanta, Dallas, Austin34
Japan
35
5.5%Tokyo, Funabashi, Kawasaki, Saitama4
Germany
32
5.0%Munich, Berlin, Stuttgart, Any city4
Canada
27
4.3%Toronto, Ottawa, Montreal, Calgary1
Switzerland
24
3.8%Bern, Geneva, Zurich, Furttal2
Spain
22
3.5%Barcelona, Madrid, Any city, A Coruna1
Italy
21
3.3%Rome, Milan, Any city, Turin1
United Kingdom
18
2.8%London, Any city, Bristol, Cambridge5
France
15
2.4%Paris, Any city, Toulouse, Valence2
Australia
14
2.2%Melbourne, Any city, Gold Coast, Sydney3
Malaysia
14
2.2%Any city, Butterworth, George Town, Ipoh2
Netherlands
14
2.2%Amsterdam, Eindhover, Rotterdam, The Hague2
Philippines
14
2.2%Angeles, Calamba, Caloocan, Cebu City1
Indonesia
12
1.9%Jakarta, Bandung, Batam, Bekasi1
Belgium
10
1.6%Brussels, Leuven1
Sweden
8
1.3%Stockholm3
Thailand
8
1.3%Bangkok, Ayutthaya, Chiang Mai, Chonburi1
Vietnam
7
1.1%Bac Ninh, Bien Hoa, Da Nang, Hai Phong1
Norway
6
0.9%Oslo, Lรธrenskog1
Austria
5
0.8%Vienna1
Bulgaria
5
0.8%Sofia, Any city1
Cambodia
5
0.8%Battambang, Phnom Penh, Siem Reap, Sihanoukville1
Poland
5
0.8%Warsaw, Gdaล„sk1
Portugal
5
0.8%Lisbon, Any city1
Romania
5
0.8%Bucharest, Any city1
Finland
4
0.6%Helsinki1
Croatia
4
0.6%Any city, Zagreb1
Hungary
4
0.6%Any city, Budapest1
Czechia
3
0.5%Any city, Prague1
Denmark
3
0.5%Copenhagen1
Estonia
3
0.5%Tallinn1
Luxembourg
3
0.5%n/a1
Chile
2
0.3%Santiago1
Cyprus
2
0.3%Nicosia1
Greece
2
0.3%Thessaloniki1
Ireland
2
0.3%Dublin1
Lithuania
2
0.3%Vilnius1
Latvia
2
0.3%Riga1
New Zealand
2
0.3%Wellington1
Saudi Arabia
2
0.3%Adelaide, Riyadh1
Singapore
2
0.3%Singapore, Singapore Office2
Slovenia
2
0.3%Ljubljana1
Slovakia
2
0.3%Bratislava1
United Arab Emirates
1
0.2%Dubai1
Brazil
1
0.2%Mato Grosso1
Colombia
1
0.2%Bogotรก1
Ecuador
1
0.2%Quito1
India
1
0.2%Bengaluru1
Peru
1
0.2%Lima1
Ukraine
1
0.2%Lviv1
South Africa
1
0.2%Cape Town1
Not stated
4
0.6%Berlin, California - SF, Detroit, Mexico City3

Base: 634 on-site or hybrid roles still open at the close of 31 Oct 2026, placed by the country named in their location.

Countries come from the posting's own location text. A role offered at sites in two countries is counted in both, which is why the shares can total more than 100%.

Each city gets its country from the same location text: the country or US state written next to it, or the single country its list mentions. Two capital letters after a city name are treated as a US state. When a city cannot be matched, it appears with no country.

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 4

Top cities for in-person robot training roles

Which cities offer the largest number of in-person robot training roles?

City or metro areaCountryOn-site rolesNaming only this city
San Francisco Bay AreaUnited States
39
3911
DallasUnited States
13
136
AtlantaUnited States
13
124
LondonUnited Kingdom
12
115
AustinUnited States
10
103
Los AngelesUnited States
10
102
TokyoJapan
10
84
MiamiUnited States
9
93
MunichGermany
9
92
ParisFrance
9
92
PittsburghUnited States
9
74
AmsterdamNetherlands
8
82
StockholmSweden
8
83
BarcelonaSpain
7
71
BerlinGermany
7
71
Customer SiteUnited States
7
71
HoustonUnited States
7
53
BrusselsBelgium
6
61
MadridSpain
6
61
PhoenixUnited States
6
63
StuttgartGermany
6
61
TorontoCanada
6
61
BernSwitzerland
5
51
ColumbusUnited States
5
51
GenevaSwitzerland
5
51

Base: 625 of the 634 on-site roles still open at the close of 31 Oct 2026 give a city; shown are the 25 named most often.

When a role can be done at any of several sites, each of those cities counts it once. The 'Naming only this city' column shows the roles tied to that one city and no other.

Cities inside one metro area are combined. Sunnyvale, Burlingame, Santa Clara and San Francisco roll up into the San Francisco Bay Area, Redmond and Bellevue into the Seattle area, and Brooklyn into New York City.

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 5

The work behind on-site robot training roles

What tasks fill the robot training roles that require being there in person?

Work, read from the titleOn-site rolesShare of on-site roles
Autonomous vehicle and truck operation (safety drivers, test operators)326
51.4%
AV & Drone OperatorsADAS Data Collection Driver13
Robot data collection and demonstrations150
23.7%
Egocentric & Wearable CaptureAI Data Collector11
Robot service, field and fleet operations60
9.5%
Robot Fleet OperationsDepot Operator/Specialist10
Robot teleoperation and piloting48
7.6%
Robot TeleoperationInternship Teleoperation & Data Labeling8
Other16
2.5%
AV & Drone OperatorsAir Operations Specialist10
Robot operation (other)16
2.5%
Robot TeleoperationAutonomous GSE Operator11
Drone and flight operations12
1.9%
AV & Drone OperatorsDrone Pilot5
Robot testing, evaluation and QA6
0.9%
Robot Task Evaluation & QAField Development Test Associate5

Base: 634 on-site or hybrid roles still open at the close of 31 Oct 2026.

Each role counts once, in the first group whose words its title contains, in the order listed in the method (vehicles first, then drones and flight, teleoperation, data collection, service and fleet work, testing, other robot operation).

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 6

On-site versus remote pay, compared inside each kind of work

Within one kind of robot training work, do in-person roles offer higher pay than remote ones?

Kind of workOn-site median (per hour)Remote median (per hour)On-site against remote
Data & Demonstration Collection
$24
8$8031-70.6%
Robot Fleet OperationsSmall sample
$22
29n/a0n/a
AV & Drone OperatorsSmall sample
$35
18n/a0n/a
Robot TeleoperationSmall sample
$26
6n/a1n/a
Egocentric & Wearable CaptureSmall sample
n/a
1$1328n/a
Embodied-AI Data AnnotationSmall sample
n/a
1$74n/a
Robot Task Evaluation & QASmall sample
n/a
1n/a0n/a

Base: Roles still open at the close of 31 Oct 2026 with an hourly rate, limited to kinds of work that have on-site roles.

A median appears only when each side has 3 or more roles with a stated rate. Comparing inside each kind of work keeps a different mix of jobs from looking like a premium.

The medians leave out on-site roles that list a yearly salary, a per-session fee or no pay figure.

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Figure 7

Like-for-like change in on-site roles across the window

Are in-person robot training roles becoming more or less common?

Work arrangementLive on 1 Oct 2026Live on 31 Oct 2026Change
On-site or hybrid568605
+6.5%
69
RemoteSmall sample2515
-40.0%
14
Not stated11394
-16.8%
92
All roles706714
+1.1%
175

Base: Open roles as 1 Oct 2026 began and as 31 Oct 2026 closed, on employers tracked the whole time, plus every new listing in the window.

Only employers already tracked before the window began enter the open counts. An employer added later cannot show up as growth.

New listings: roles first seen in the window on an employer already tracked before that week (an employer's first week is its back catalogue).

A row with under 36.8 open roles at the start carries a small-sample mark and is left unranked: under that count, the average across all rows says more than the row's own open roles at the start do (a weight fitted on the counts themselves).

Credit: RobotTraining.jobs Research, On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition.

Robotics labs ยท Fleet operators ยท Data vendors ยท Research teams

Buy the data behind this report

This report sums up records we gather ourselves, day after day:

  • 1,432 robot training roles followed, history included

    Every role on 100 employer job boards, read each day since 12 Sep 2026: its pay, its place, and the first and last day it was up.

  • 536 apply clicks over 28 days

    Which robot training roles people really apply to, by kind of work, pay and place. The clicks happen here, so no one else can count them.

  • 25 sourced industry datasets

    Funding, revenue, deals, fleets, deployments and prices across the robot training industry, each record linked to the publication it comes from.

On offer

  • The raw records

    Every robot training role we track: title, employer, place, kind of work, pay as stated, and the days it was first and last open.

  • A regular feed

    New and closed roles every week or month, as file drops or through the data endpoints.

  • A custom cut

    Any figure here by country, city, kind of work, vehicle, employer or date, or your own hiring benchmarked against the market.

Need the people behind this data for your own robots? (you contract them yourself) or (we deliver the robot data).

Investing in robotics, AV or robot-data companies? The Robot hiring radar is these records role by role, every day, as files.

Robot hiring brief: free, once a month

The robot-training job market in figures, on the first weekday of each month: open roles, what opened and closed, who added and who cut, and posted pay by kind of work. To your work email.

How it was measured

Daily reads of the public job boards run by the robot training employers and programs this site follows, using each open role's title, location text and the workplace field its board provides. A single database query produces each figure for the window shown, and the live page and every monthly edition run that identical query. The note under each figure gives its base.

On-site or hybrid
The work happens at a named location for at least part of the week. Either the title or location says so ('Onsite', 'On-site London', 'In-person', 'Hybrid'), or a board with its own workplace field (Lever, Ashby) marks the role as not remote. Hybrid roles land in this group because listing fields cannot tell them apart from fully on-site ones.
Remote
The role's title or location says remote (or work from home, worldwide), or its board flags it remote. 'Remote' that names the work ('Remote Driver', 'Remote Assistance Operator') says how a robot or vehicle is controlled, not where the worker sits, and is not read as remote.
Not stated
Nothing in the title, location or workplace field indicates where the work happens. Silence is never taken to mean on-site.
City or metro area
Taken from the location text. A city's country comes from the country or US state written next to it, or from the only country in its list, and is never guessed. Cities in the same metro area are combined, and a role offered at several sites counts once at each.
Work, read from the title
Each on-site role falls in the first of these groups whose words its title contains: autonomous vehicle and truck operation; drone and flight operations; robot teleoperation; robot data collection and demonstrations; robot service, field and fleet operations; robot testing, evaluation and QA; other robot operation; other.
Advertised pay
The hourly rate a listing gives (midpoint of any range), compared inside each kind of work. A median needs 3 or more roles with a stated rate on each side.
New listing
A role seen for the first time during the window, at an employer this site was already following the week before. Whatever an employer has up in its first tracked week is back catalogue and does not count as new.
Open role
On a given day, a role that has been posted and still shows in the employer's feed.
Window
Live page: the 28 most recent full days (UTC). Edition: a single calendar month (UTC).

Limits of this data

  • Both shares are floors: a role that requires on-site sessions may say so only in its description.
  • Listing fields give no way to separate hybrid roles from fully on-site ones.
  • Location text can mean where the work happens or where applicants must live, and it is not always clear which. Two capital letters after a city are taken as a US state.
  • Hiring that happens without a public job board is not captured.
  • A posting is not a hire, and employers that post one copy per city or site give extra weight to the cities where they do that a lot.

Collection and classification of listings are explained on the methodology page. Freshness, completeness and consistency are tested daily in the data quality report.

Monthly editions

The live page spans the 28 most recent full days and updates daily. Every finished month also gets its own edition, locked two days after month end, so a number quoted from it stays fixed.

  • Live page28 most recent full days, rebuilt daily
  • October 2026 edition (you are here)Will lock when the next daily refresh runs

Quoting this report

You may quote, chart or republish these figures under CC BY 4.0, provided you link back to this page. This edition is locked: its numbers will not change.

Citation

RobotTraining.jobs (2026). On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition. https://robottraining.jobs/research/on-site-robot-training-work/2026-10

BibTeX

@misc{robottrainingjobs_on_site_robot_training_work_2026_10,
  author = {{RobotTraining.jobs}},
  title = {On-Site Robot Training Jobs: Labs, Depots and Test Tracks, October 2026 edition},
  year = {2026},
  howpublished = {\url{https://robottraining.jobs/research/on-site-robot-training-work/2026-10}},
}
Files: every figure in one CSV (a row per value) or as JSON, emailed to your work address.Other open data: datasets and API

Keep reading

Browse all 91 reports