Update app.py

app.py

@@ -503,6 +503,12 @@ if apply_filters:
     # Filter Group Model
     if filter_dict.get('group_model') is not None:
         df = df[df['group_model'] == filter_dict['group_model']]
+    
+    # UI display mapping (for rendering only — data remains unchanged)
+    group_model_display = {
+        'OB HAULLER': 'OB HAULER',
+        'HAULING COAL': 'COAL HAULING'
+    }
 
     # Filter Shift
     if filter_dict.get('shift') is not None:
@@ -1760,7 +1766,6 @@ with col_insights:
 
     # ===================== 2. High-Speed Fatigue Analysis =====================
     if col_speed and col_speed in df.columns:
-
         high_speed_threshold = 20
         high_speed_fatigue = df[df[col_speed] >= high_speed_threshold]
         high_speed_pct = (len(high_speed_fatigue) / len(df)) * 100 if len(df) > 0 else 0
@@ -1783,13 +1788,11 @@ with col_insights:
             st.info(
                 f"{high_speed_pct:.1f}% of alerts occur at high speeds. This is within acceptable range."
             )
-
     else:
         st.info("Speed data not available for High-Speed Fatigue Analysis.")
 
     # ===================== 3. Shift Pattern Analysis =====================
     if col_shift and col_shift in df.columns:
-
         shift_counts = df[col_shift].value_counts()
         st.markdown(f"**Shift Pattern Risk**")
 
@@ -1813,13 +1816,11 @@ with col_insights:
                 st.info(
                     f"Shift {shift_val} alert distribution is acceptable ({shift_pct:.1f}%)."
                 )
-
     else:
         st.info("Shift data not available for Shift Pattern Analysis.")
 
     # ===================== 4. Operator Risk Profiling =====================
     if col_operator and col_operator in df.columns:
-
         operator_alerts = df[col_operator].value_counts()
         top_risk_operators = operator_alerts.head(5)
 
@@ -1848,129 +1849,148 @@ with col_insights:
                 st.info(
                     f"Operator {op_name} fatigue risk is within acceptable range ({op_pct:.1f}%)."
                 )
-
     else:
         st.info("Operator data not available for Operator Risk Profiling.")
 
-
 # =====================================================================
-# 🔹 KOLOM KANAN — AI RECOMMENDATIONS
+# 🔹 KOLOM KANAN — AI RECOMMENDATIONS (PER INSIGHT)
 # =====================================================================
 with col_recs:
-
     st.subheader("Recommendations")
-    ai_recs = []
-    insights_found = []
 
-    # Peak Hour
+    # Reset list to collect recommendations per insight
+    ai_recommendations = []
+
+    # 1. Critical Hour Insight → AI Rec
     if "hour" in df.columns and not df.empty:
         peak_hour = df["hour"].value_counts().idxmax()
         critical_hours = [2, 3, 4, 5]
 
         if peak_hour in critical_hours:
-            insights_found.append(
-                f"Most fatigue risk occurs at **{peak_hour}:00** — during critical circadian low period (3-6 AM)."
-            )
+            ai_recommendations.append({
+                "action": "Deploy enhanced fatigue monitoring systems during 3-6 AM.",
+                "data_point": f"Critical Hour Alerts: {len(critical_alerts)} ({critical_pct:.1f}%)",
+                "reasoning": "High percentage of alerts during circadian low period."
+            })
         else:
-            insights_found.append(
-                f"Most fatigue risk occurs at **{peak_hour}:00** — likely due to circadian drop."
-            )
-
-    # Risk Shift
-    if col_shift and col_shift in df.columns and not df.empty:
-        worst_shift = df[col_shift].value_counts().idxmax()
-        insights_found.append(
-            f"Highest fatigue recorded in **Shift {worst_shift}** — review scheduling & workload."
-        )
-
-    # Worst Operator
-    if col_operator and col_operator in df.columns and not df.empty:
-        worst_operator = df[col_operator].value_counts().idxmax()
-        insights_found.append(
-            f"Operator at highest risk: **{worst_operator}** — suggested coaching or rest plan."
-        )
-
-    # Duration Risk
-    if "duration_sec" in df.columns and not df.empty:
-        avg_duration = df["duration_sec"].mean()
-        if avg_duration > 10:
-            insights_found.append(
-                "Long fatigue event duration suggests slow response — improve alerting training."
-            )
+            ai_recommendations.append({
+                "action": "Monitor fatigue patterns around peak hour (Hour {peak_hour}).",
+                "data_point": f"Peak Hour: {peak_hour}:00 — {df['hour'].value_counts()[peak_hour]} alerts",
+                "reasoning": "This hour shows highest fatigue occurrence."
+            })
 
-    # ===================== AI DECISION ENGINE =====================
-    if insights_found:
+    # 2. High-Speed Insight → AI Rec
+    if col_speed and col_speed in df.columns and not df.empty:
+        high_speed_threshold = 20
+        high_speed_fatigue = df[df[col_speed] >= high_speed_threshold]
+        high_speed_pct = (len(high_speed_fatigue) / len(df)) * 100 if len(df) > 0 else 0
 
-        if any("circadian" in i.lower() for i in insights_found):
-            ai_recs.append({
-                "recommendation": "Deploy enhanced fatigue monitoring systems during 3-6 AM.",
-                "data_point": f"Critical Hour Alerts: {len(critical_alerts)} ({critical_pct:.1f}%)",
-                "reason": "High percentage of alerts during circadian low period."
+        if high_speed_pct > 20:
+            ai_recommendations.append({
+                "action": "Implement speed-reduction protocols during fatigue-prone hours.",
+                "data_point": f"High-Speed Alerts: {len(high_speed_fatigue)} ({high_speed_pct:.1f}%)",
+                "reasoning": "High-speed alerts increase accident severity potential."
             })
-
-        if any("shift" in i.lower() for i in insights_found):
-            ai_recs.append({
-                "recommendation": "Review shift rotation schedules.",
-                "data_point": f"Shift {worst_shift}: {df[col_shift].value_counts()[worst_shift]} alerts",
-                "reason": "This shift shows highest fatigue alerts."
+        else:
+            ai_recommendations.append({
+                "action": "Maintain current speed monitoring — risk level is acceptable.",
+                "data_point": f"High-Speed Alerts: {len(high_speed_fatigue)} ({high_speed_pct:.1f}%)",
+                "reasoning": "Current high-speed fatigue rate is within acceptable range."
             })
 
-        if any("operator" in i.lower() for i in insights_found):
-            ai_recs.append({
-                "recommendation": "Coaching or mandatory rest for the identified high-risk operator.",
-                "data_point": f"Operator {worst_operator}: {df[col_operator].value_counts()[worst_operator]} alerts",
-                "reason": "Operator has highest fatigue alerts."
-            })
+    # 3. Shift Pattern Insight → AI Rec
+    if col_shift and col_shift in df.columns and not df.empty:
+        worst_shift = df[col_shift].value_counts().idxmax()
+        shift_pct = (df[col_shift].value_counts()[worst_shift] / len(df)) * 100
 
-        if any("duration" in i.lower() for i in insights_found):
-            ai_recs.append({
-                "recommendation": "Improve fatigue alert response training.",
-                "data_point": f"Avg Duration: {avg_duration:.1f} sec",
-                "reason": "Long fatigue event duration indicates slow response."
+        if shift_pct > 50:
+            ai_recommendations.append({
+                "action": "Review shift rotation schedules for Shift {worst_shift}.",
+                "data_point": f"Shift {worst_shift}: {df[col_shift].value_counts()[worst_shift]} alerts ({shift_pct:.1f}%)",
+                "reasoning": "Disproportionately high fatigue alerts indicate scheduling imbalance."
+            })
+        else:
+            ai_recommendations.append({
+                "action": "Continue monitoring all shifts — no dominant risk identified.",
+                "data_point": f"Shift {worst_shift}: {df[col_shift].value_counts()[worst_shift]} alerts ({shift_pct:.1f}%)",
+                "reasoning": "Shift distribution is balanced."
             })
 
-        # Render all recommendations
-        import re
+    # 4. Operator Risk Insight → AI Rec
+    if col_operator and col_operator in df.columns and not df.empty:
+        worst_operator = df[col_operator].value_counts().idxmax()
+        op_pct = (df[col_operator].value_counts()[worst_operator] / len(df)) * 100
 
-        for rec in ai_recs:
+        if op_pct > 5:
+            ai_recommendations.append({
+                "action": "Coaching or mandatory rest for the identified high-risk operator.",
+                "data_point": f"Operator {worst_operator}: {df[col_operator].value_counts()[worst_operator]} alerts ({op_pct:.1f}%)",
+                "reasoning": "Operator has highest fatigue alerts — requires individual intervention."
+            })
+        else:
+            ai_recommendations.append({
+                "action": "Continue general monitoring — no single operator dominates risk.",
+                "data_point": f"Top Operator: {worst_operator} — {df[col_operator].value_counts()[worst_operator]} alerts ({op_pct:.1f}%)",
+                "reasoning": "Risk is distributed across operators — no urgent individual action needed."
+            })
 
-            data_point_colored = re.sub(
-                r'(\d+\.?\d*%)',
-                r'<span style="color: red;">\1</span>',
-                rec['data_point']
-            )
+    # Render each recommendation as a card
+    for rec in ai_recommendations:
+        # Highlight percentages in red
+        data_point_colored = rec['data_point'].replace(
+            f"({rec['data_point'].split('(')[-1]}",
+            f"(<span style='color: red;'>{rec['data_point'].split('(')[-1]}"
+        ).replace(")", "</span>)")
 
-            reason_colored = re.sub(
-                r'(\d+\.?\d*%)',
-                r'<span style="color: red;">\1</span>',
-                rec['reason']
-            )
+        reasoning_colored = rec['reasoning'].replace(
+            f"({rec['reasoning'].split('(')[-1]}",
+            f"(<span style='color: red;'>{rec['reasoning'].split('(')[-1]}"
+        ).replace(")", "</span>)")
 
-            st.markdown(
-                f"""
+        st.markdown(
+            f"""
+            <div style="
+                background: #f8f9fa;
+                border: 1px solid #dee2e6;
+                border-radius: 8px;
+                padding: 15px;
+                margin: 10px 0;
+                box-shadow: 0 2px 8px rgba(0,0,0,0.05);
+            ">
                 <div style="
-                    background: #f8f9fa;
-                    border: 1px solid #dee2e6;
-                    border-radius: 8px;
-                    padding: 15px;
-                    margin: 10px 0;
+                    font-weight: bold;
+                    background: #e9ecef;
+                    padding: 8px;
+                    border-radius: 5px;
+                    margin-bottom: 8px;
+                    border-left: 4px solid #495057;
                 ">
-                    <div style="font-weight: bold; background: #e9ecef; padding: 8px; border-radius: 5px;">
-                        AI Recommendation
-                    </div>
-                    <div style="padding-top: 8px;"><strong>Action:</strong> {rec['recommendation']}</div>
-                    <div style="padding: 8px; background: #e9ecef; border-radius: 5px;">
-                        <strong>Data Point:</strong> {data_point_colored}
-                    </div>
-                    <div style="padding: 8px; background: #f1f1f1; border-radius: 5px;">
-                        <strong>AI Reasoning:</strong> {reason_colored}
-                    </div>
+                    AI Recommendation
                 </div>
-                """,
-                unsafe_allow_html=True
-            )
+                <div style="padding: 8px 0;">
+                    <strong>Action:</strong> {rec['action']}
+                </div>
+                <div style="
+                    padding: 8px;
+                    background: #f1f1f1;
+                    border-radius: 5px;
+                    margin: 8px 0;
+                ">
+                    <strong>Data Point:</strong> {data_point_colored}
+                </div>
+                <div style="
+                    padding: 8px;
+                    background: #f1f1f1;
+                    border-radius: 5px;
+                ">
+                    <strong>AI Reasoning:</strong> {reasoning_colored}
+                </div>
+            </div>
+            """,
+            unsafe_allow_html=True
+        )
 
-    else:
+    if not ai_recommendations:
         st.info(
             "No specific data points available for AI recommendations. "
             "Ensure relevant columns are present (hour, shift, operator, duration, speed)."
@@ -1981,4 +2001,4 @@ st.markdown("---")
 st.markdown(
     '<div class="footer">FatigueAnalyzer - Transforming Mining Safety with Intelligent Analytics | Contact: info@bukittechnology.com</div>',
     unsafe_allow_html=True
-)
+)