BRENNEN

Industries / Health / Healthcare / Home health

Scheduling and routing visits

Deciding which nurse, therapist or aide visits which home, at what time and in what order, and recording that each visit happened.

7

Milestones achieved, from 1997 to 2024.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1997

    Map software plans nurse routes

    The University of Alabama's Productivity Center and a Visiting Nurses Association built a planning system that combined PC map software with scheduling rules, working within doctors' limits on weekly visits per patient. The authors report it saved the association over $20,000 a year in travel and planning time and spread work more evenly among nurses.

    ideas.repec.org

    Credit

    Sachidanand V. Begur, David M. Miller and Jerry R. Weaver (Interfaces, 1997)

  2. 2006

    City-wide planning in Sweden

    Laps Care is a planning system built on operations research, the mathematics of scheduling and routing, that makes daily visit plans for home care staff. INFORMS reports about 200 units and 4,000 staff in 50 Swedish municipalities using it; Stockholm, a user since 2003, cut planning time by about two-thirds and raised efficiency by 10–15%. These figures come from the system's makers.

    informs.org

    Credit

    Patrik Eveborn, Patrik Flisberg and Mikael Rönnqvist (European Journal of Operational Research, 2006); City of Stockholm

  3. 2014

    A shared test set

    The authors published a model of a day of home visits with two hard cases: one service must follow another after a set gap, such as medicine before a meal, and two staff must arrive at once, for example to lift a patient. They released their test problems as public files, which later groups used to compare methods; Kummer (2021) added 17,600 more built on real addresses in Porto Alegre, Brazil.

    econpapers.repec.orgprodlog.wiwi.uni-halle.de

    Credit

    Dorota Mankowska, Frank Meisel and Christian Bierwirth (Health Care Management Science, 2014)

  4. 2017

    Nurses set their own rotas

    A London pilot of the Dutch Buurtzorg model ran from January to July 2017. The nurse team set its own weekly rotas in team meetings and agreed time off among themselves. An independent study found the model feasible and liked by patients, but nurses found self-management hard to learn and some left.

    Lessonself-scheduling teams needed a coach and back-office support in place before they started.

    pmc.ncbi.nlm.nih.gov

    Credit

    Vari Drennan and colleagues (BMJ Open, 2018), evaluating the model created by Jos de Blok

  5. 2020

    Phone check-ins become law

    Electronic visit verification (EVV) means a phone or device records when and where a Medicaid-paid home visit happened. Section 12006 of the US 21st Century Cures Act made it mandatory from 1 January 2020 for personal care and 1 January 2023 for home health; states that miss the deadline can lose up to 1% of their federal Medicaid matching rate.

    medicaid.gov

    Credit

    US Congress; Centers for Medicare & Medicaid Services (CMS)

  6. 2021

    EVV's hidden costs measured

    An interview study of workers, service users and advocates found that EVV added unpaid work for caregivers, and that because the app tracks the worker, it also tracked the movements of the disabled and older people they served.

    Lessona checking tool designed without watching how home care is actually done can cut into the care it is meant to protect.

    datasociety.net

    Credit

    Alexandra Mateescu, Data & Society

  7. 2024

    Test problems on real roads

    A new generator places patients according to the EU's 1 km population grid and computes car travel times on real roads with OSRM, the Open Source Routing Machine. It produced 200 test problems across urban, rural and mountain areas of Italy, with a common file format and a validator; the authors' own method beat the earlier published methods on all of them.

    air.uniud.it

    Credit

    Sara Ceschia, Luca Di Gaspero, Roberto Maria Rosati and Andrea Schaerf, University of Udine (International Transactions in Operational Research, 2024)

Watch for

  1. An agent refills cancelled visits

    When a caregiver calls off, AlayaCare's Vacant Visit Scheduling Agent, a software agent, finds replacements, sends shift offers, tells clients and passes on what it cannot solve. The company's webinar slides claim "70%+ disruptions resolved without manual effort".

    Achieved whenan agency or independent evaluator publishes fill rates and continuity of care from real use, compared with human schedulers.

    alayacare.com

    Claimed by

    AlayaCare, March 2026

  2. An agent fixes failed check-ins

    The same slides claim AlayaCare's Visit Verification Agent resolves "80%+ EVV failures" by analysing visit data and messaging caregivers.

    Achieved whena state Medicaid agency or independent auditor reports error and resolution rates for agencies that use it.

    alayacare.com

    Claimed by

    AlayaCare, March 2026

Open

  1. 01

    A multi-day planning benchmark

    A public set of synthetic home-visit planning problems that cover 5 to 7 days instead of one, scored on how often each patient sees the same caregiver (continuity of care) as well as on travel time and lateness, with real road travel times in at least three countries outside Italy.

    The brief
    Credit · unclaimed
  2. 02

    Open solvers on shared tests

    A reproducible run of general-purpose open-source routing solvers (OR-Tools, Timefold, PyVRP, VROOM) on the Mankowska, Kummer and Udine test sets, reporting each one's gap to the best-known answer and which of them can express two-person visits and timed gaps at all.

    The brief
    Credit · unclaimed

Open 01 · the brief

A multi-day planning benchmark

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public set of synthetic home-visit planning problems that cover 5 to 7 days instead of one, scored on how often each patient sees the same caregiver (continuity of care) as well as on travel time and lateness, with real road travel times in at least three countries outside Italy.

Why it is still open

The Udine authors list a multi-day horizon and caregiver stability as future work, and their benchmark repository, archived in November 2025, holds no multi-day files. Searches found no public multi-day set with real roads.

github.com

Built with what exists

the EU GEOSTAT population grid or WorldPop for where people live; OpenStreetMap and OSRM for travel times; the Udine JSON format and Python validator, extended to several days; Google OR-Tools, Timefold or PyVRP as baseline solvers.

  1. First step

    Write a two-day extension of the Udine JSON format and generate 10 problems for one non-Italian city with OSRM.

  2. Done looks like

    At least 100 problems published with a DOI, for example on Zenodo, covering three or more countries, with a validator that checks continuity scores and baseline results from at least two open-source solvers, with run times.

Open 02 · the brief

Open solvers on shared tests

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A reproducible run of general-purpose open-source routing solvers (OR-Tools, Timefold, PyVRP, VROOM) on the Mankowska, Kummer and Udine test sets, reporting each one's gap to the best-known answer and which of them can express two-person visits and timed gaps at all.

Why it is still open

The Udine paper compares only research methods. Searches found no published comparison of these widely used open-source tools on these sets, so agencies that build on them have no public yardstick.

air.uniud.it

Built with what exists

Python; OR-Tools; Timefold Solver; PyVRP; VROOM; the Udine validator to check every answer.

  1. First step

    Run OR-Tools on the 10 smallest Udine problems and check the answers with the validator.

  2. Done looks like

    A public repository with scripts and a results table covering all 200 Udine problems and the Mankowska set, every answer passing the validator, with cost, gap to best-known and time limit for each run, and a list of the constraints each solver could not model.