SITUATIONAL AWARENESS FOR FIELD TEAMS
The conditions change. Stay a step ahead.
Site risks, crew alerts, and tomorrow’s outlook.
All in one place.
One secure sign-in. The right workspace for your role.
HOW ForeSite WORKS
ForeSite brings environmental conditions, on-site observations, and a next-day risk outlook into one picture for each job site.
About this preview Plant locations are real; conditions and scores are sample data. The model design below follows our scoring plan, but this preview does not run trained models. Live alerts, team-wide report sharing, and briefing delivery require connected services.
Select a real power plant to review weather, extreme heat, lightning, flooding, wildfire, soil saturation, and wildlife. Read each source and timestamp; missing readings show “no data.”
A lightning strike within 10 miles triggers a stop-work alert and a 30-minute countdown. Each new nearby strike restarts it. River sites can also receive flood warnings after heavy overnight rain.
Add a photo, pin the location, and describe a hazard for your team and future crews. An open high-severity report creates a warning until a supervisor resolves it.
Supervisors compare sites on a map ranked by hazard score, inspect the factors behind each score, and review crew observations before allocating attention and resources.
At 6 PM in each site’s local timezone, the daily briefing summarizes tomorrow’s hazards using the forecast and score. Open the briefing to review its date and the information it was based on.
HAZARD SCORE / 0–10 · PROTOTYPE
The index estimates whether selected National Weather Service warnings or heat advisories may cover a plant during the next 24-hour forecast window. It is not an injury prediction or a guarantee that other hazards are absent.
One model was selected from five candidates using fixed-lead GFS forecasts and archived warning areas, with separate county and time evaluations.
Readable features such as 24-hour rain, max wind, max heat index and seasonal patterns explain the score. A feature can relate to more than one hazard.
Prototype · model estimate × 10
Dated GBIF sightings near a plant show the species, source, distance and available coordinate uncertainty.
Sightings and crew reports do not contribute score points or establish an encounter probability. Missing sightings do not mean wildlife is absent.
The baseline is the model's starting value. Signed feature contributions raise or lower it to reach the final index. For example: 2.1 baseline + 2.6 rain + 1.2 wind + 0.9 heat index − 0.4 seasonal contribution = 6.4. These are illustrative values, not current readings.
Prototype status stays visible. The current research model missed one of its evaluation checks. Scores remain labeled Prototype and show their target window; they must not be presented as established probabilities.
Freshness matters. Inputs older than two hours, expired forecasts and invalid or unapproved predictions show no data. A missing score is never zero.
Alerts are independent. The index never clears lightning stop-work instructions, flood warnings or crew-report warnings. A red-flag warning means fire-weather conditions; a severe-storm warning does not confirm a measured lightning strike.
TAKE A LOOK AROUND
Explore crew and supervisor workflows.
Sample data is clearly labeled in each workspace.