diff --git a/entries/powerplant_bayesian/__init__.py b/entries/powerplant_bayesian/__init__.py
index 27361d2..9e2855a 100644
--- a/entries/powerplant_bayesian/__init__.py
+++ b/entries/powerplant_bayesian/__init__.py
@@ -21,7 +21,7 @@ from .. import *
 import re
 
 title = "Power plant sensor Bayesian network (2007)"
-description = """Bayesian network modelling the correlations among the sensors of a coal-fired power plant, from T. D. Nielsen and F. V. Jensen, distributed with the book "Bayesian Networks and Decision Graphs" (F. V. Jensen and T. D. Nielsen) at Aalborg University. The network was used to design an on-line alert system that detects abnormal plant behaviour as conflicts between sensor readings and the model. Nodes are variables (sensor measurements, discretized into three intervals), and a directed edge from $i$ to $j$ means that variable $i$ is a parent of variable $j$ in the directed acyclic graph of the Bayesian network, i.e., that the conditional probability distribution of $j$ is specified given the value of $i$. The network has 46 sparsely connected nodes (six of which have no connections); the conditional probability tables are not included.
+description = """Bayesian network modelling the correlations among the sensors of a coal-fired power plant, from T. D. Nielsen and F. V. Jensen, distributed with the book "Bayesian Networks and Decision Graphs" (F. V. Jensen and T. D. Nielsen) at Aalborg University. The network was used to design an on-line alert system that detects abnormal plant behavior as conflicts between sensor readings and the model. Nodes are variables (sensor measurements, discretized into three intervals), and a directed edge from $i$ to $j$ means that variable $i$ is a parent of variable $j$ in the directed acyclic graph of the Bayesian network, i.e., that the conditional probability distribution of $j$ is specified given the value of $i$. The network has 46 sparsely connected nodes (six of which have no connections); the conditional probability tables are not included.
 
 The nodes have the following properties:
 

diff --git a/entries/powerplant_bayesian/__init__.py b/entries/powerplant_bayesian/__init__.py
new file mode 100644
index 0000000..27361d2
--- /dev/null
+++ b/entries/powerplant_bayesian/__init__.py
@@ -0,0 +1,74 @@
+#! /usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# Copyright (C) 2020-2026 Tiago de Paula Peixoto <tiago@skewed.de>
+#
+# This program is free software: you can redistribute it and/or modify it under
+# the terms of the GNU Affero General Public License as published by the Free
+# Software Foundation, either version 3 of the License, or (at your option) any
+# later version.
+#
+# This program is distributed in the hope that it will be useful, but WITHOUT
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+# FOR A PARTICULAR PURPOSE.  See the GNU Affero General Public License for more
+# details.
+#
+# You should have received a copy of the GNU Affero General Public License along
+# with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+
+from .. import *
+import re
+
+title = "Power plant sensor Bayesian network (2007)"
+description = """Bayesian network modelling the correlations among the sensors of a coal-fired power plant, from T. D. Nielsen and F. V. Jensen, distributed with the book "Bayesian Networks and Decision Graphs" (F. V. Jensen and T. D. Nielsen) at Aalborg University. The network was used to design an on-line alert system that detects abnormal plant behaviour as conflicts between sensor readings and the model. Nodes are variables (sensor measurements, discretized into three intervals), and a directed edge from $i$ to $j$ means that variable $i$ is a parent of variable $j$ in the directed acyclic graph of the Bayesian network, i.e., that the conditional probability distribution of $j$ is specified given the value of $i$. The network has 46 sparsely connected nodes (six of which have no connections); the conditional probability tables are not included.
+
+The nodes have the following properties:
+
+- `name`: the name of the variable, as in the source (e.g. "V26"; the variables are not described further).
+- `states`: the list of states of the variable (intervals of the sensor value, e.g. "(2.42-4.66]").[^icon]
+
+[^icon]: Description obtained in part from the [ICON](https://icon.colorado.edu) project."""
+tags = ['Informational', 'Unweighted', 'Metadata']
+url = 'https://web.archive.org/web/20070626181030/http://bndg.cs.aau.dk/html/bayesian_networks.html'
+citation = [('T. D. Nielsen, F. V. Jensen, "On-line alert systems for production plants: A conflict based approach." International Journal of Approximate Reasoning 45(2), 255-270 (2007)', 'https://doi.org/10.1016/j.ijar.2006.06.010')]
+icon_hash = '57d856c967351c5e0ccda0c9'
+upstream_license = None
+# The book's site (bndg.cs.aau.dk) is no longer reachable; the file is
+# obtained from the Internet Archive.
+upstream_prefix = 'https://web.archive.org/web/20070626181753id_/http://bndg.cs.aau.dk/Bayesian_networks'
+files = [('powerplant.net', None, None)]
+
+def fetch_upstream(force=False):
+    return retry_download(fetch_upstream_files, __name__.split(".")[-1],
+                          upstream_prefix, files, force)
+
+@cache_network()
+@coerce_props()
+@annotate()
+def parse(alts=None):
+    global files
+    name = __name__.split(".")[-1]
+    for fname, alt, fmt in files:
+        if alts is not None and alt not in alts:
+            continue
+        with open_upstream_file(name, fname, "r") as f:
+            text = f.read()
+        # Hugin network specification: "node X { ... states = (...); ... }"
+        # and "potential (X | parents) { ... }"
+        nodes = {}
+        for m in re.finditer(r"^node\s+(\w+)\s*\{(.*?)\}", text, re.S | re.M):
+            st = re.search(r"states\s*=\s*\((.*?)\);", m.group(2), re.S)
+            nodes[m.group(1)] = re.findall(r'"([^"]*)"', st.group(1)) if st else []
+        idx = {v: i for i, v in enumerate(nodes)}
+        edges = []
+        for m in re.finditer(r"^potential\s*\(\s*(\w+)\s*(?:\|([^)]*))?\)",
+                             text, re.M):
+            for p in (m.group(2) or "").split():
+                edges.append((idx[p], idx[m.group(1)]))
+        g = Graph(directed=True)
+        g.add_vertex(len(nodes))
+        g.vp.name = g.new_vp("string", vals=list(nodes.keys()))
+        g.vp.states = g.new_vp("vector<string>", vals=list(nodes.values()))
+        g.add_edge_list(edges)
+        yield alt, g