Scope: Current October 2026 preview. The earlier Beta editions may have different buttons, saved data locations or included features. Always back up data before testing an updated Windows installer.
1. Purpose and operating promise
Work Study Lab is not a stopwatch-only productivity score. It helps a manager define a real problem, observe how work happens, record representative evidence, distinguish necessary work from avoidable delay, test proposed changes, and check safety and quality before deciding on a standard.
2. Install and protect existing work
Use the published Windows preview .exe or ZIP. This preview is unsigned. Earlier app profiles are not guaranteed to migrate; export complete study packages and preserve previous local backups before making changes. Hosted voice, document processing and optional phone/tablet connections require internet access. You can perform core local observations without those hosted services.
3. Choose a work mode
Guided gives step-by-step Gravity assistance; Assisted provides more manager control with available support; Independent supports experienced users; Passive reduces interruptions. Choose the mode appropriate to the observer. A manager owns the decisions and verification regardless of mode.
4. Critical Guided startup instruction
Click Guided once, then wait for Gravity to join you before clicking anything else. Do not repeatedly click while expert layers initialize. A loading pause is not an instruction to keep pressing the button. If the screen becomes unresponsive, preserve your saved work and restart the program; report the action and last visible screen.
5. Start with a measurable problem
Give the study a title and location/station. State the observed problem in neutral language, for example “During Friday lunch, pizza assembly waits for finished dough.” Record suspected causes separately so the hypothesis is not mistaken for proof. Specify what better performance would look like without compromising food safety or quality.
6. Set a clear study boundary
Define precisely where the operation starts and ends: for example a ticket arriving on the pizza rail through the plated pizza at pickup. Record roles, equipment, conditions, demand/source, shifts and employee experience. An unclear boundary makes before/after results unreliable.
7. Choose observation and timing method
Use the appropriate workflow for task timing, motion/travel, bottleneck/queue, prep batch, dish-room flow, line setup, service cycle, staffing workload or work sampling. Select active timing or manual observation as appropriate. Measure the entire work and all important elements, not only the fastest visible motion.
8. Describe work elements before timing
Break the process into understandable elements such as retrieve pan, assemble, cook, plate and hand off. Use consistent labels for WORK, WALK, WAIT, SEARCH, REPLENISH, REWORK or required waiting. Do not treat food-safety cleaning, temperature holds or necessary inspection as automatically removable waste.
9. Capture representative evidence
Observe enough real cycles to show differences in demand, worker proficiency, order complexity and normal operating conditions. Record interruptions and abnormal conditions. Use more than one run for standards: a single quick cycle or best employee does not establish a reliable restaurant standard.
10. Interpret demand and capacity
Enter actual demand evidence where available and keep forecast or estimated demand labeled as estimates. Use measured process cycle times, labor occupancy, batch sizes and queue behavior to assess a possible bottleneck. Not all waiting is caused by the measured employee.
11. Standard minutes, variance and allowances
Standard minutes require a defendable sample, consistent units, and review of fixed vs variable time, batch size, proficiency, learning curve and appropriate staffing allowances. Do not apply invented universal rest or personal allowances. Reject a target if it would sacrifice safe or sustainable pace.
12. Compare baseline and proposed improvement
Capture the BEFORE baseline with conditions and provenance. Describe a specific change, such as relocating utensils or changing a prep handoff. Repeat observations under comparable demand, staff and equipment conditions. Compare throughput, time, walking, error/rework and safety together; do not claim improvement only because one timing was shorter.
13. Gravity findings and human approval
Gravity can identify missing evidence, inconsistencies and priority questions through the Expert Review and triage layers. Treat each result as a recommendation tied to recorded evidence. A manager verifies the change in the restaurant and decides whether a procedure, layout or staffing change should be installed.
14. Cedar spoken guidance
Guided may use the hosted Cedar voice through an HTTPS connection. Check speakers, system volume, audio permissions and internet connectivity. Silence does not mean the study record was lost; continue with on-screen guidance. Do not depend on voice for essential safety alerts.
15. Portable phone/tablet observations
Open or save a study on Windows, then choose Connect a tablet or phone. Create a 12-hour temporary connection and open its code/link on the portable device. Record work on the floor and use Bring Back Latest when an updated portable version is available. Keep both versions protected and review possible conflicts; Windows does not silently overwrite divergent evidence.
16. Save, import and export complete evidence
The Windows program preserves local study files and durable complete-evidence packages. Prefer complete-package export when moving a study because it can include expert sidecar information beyond the base study. Confirm a package reopens before deleting anything. Keep dated copies in another secure folder or backup system.
17. Print and share a report
Use the Study Report, PDF export or Windows printing after Gravity review. Check that the executive summary, observation boundary, measured data, limitations, recommendations and verification status are correct. A printed report should distinguish evidence, interpretation and management decision.
18. Daily operator checklist
Before studying: obtain appropriate workplace permission, define scope and purpose, protect privacy, ensure safe observation and choose representative periods. During studying: note actual conditions and complete cycles. After studying: save, review discrepancies, export the full evidence package, discuss findings with staff appropriately and verify results before installing a new standard.
19. Troubleshooting
Guided freezes: stop repeated clicks, reopen and check saved evidence. Cedar silent: confirm internet/audio and rely on text while troubleshooting. Portable device not connecting: confirm both devices are online, the session is valid, and the link/code are correct. Device edits missing: check Bring Back Latest and conflict indicators. DOCX extraction fails: internet is required for hosted DOCX processing; paste text or use a supported local text file instead.
20. Glossary
Study boundary = exact start and end of observed work. Work element = distinct observable task. Baseline = before-change measurement with recorded conditions. Representative sample = data reflecting normal variations. Standard minute = defensible time allowance based on valid evidence. Portable connection = time-limited device session. Evidence package = saved base study plus supporting expert observations.